CD Ma Student Nortel

CDMA

Technology OverviewStudent GuideNBSS9.0 Standard 04.02 October 2000

Course 809A

CDMA

Technology OverviewStudent Guide

Course number: Course 809AProduct release: NBSS9.0Document version: Standard 04.02Date: October 2000

Copyright Country of printing Confidentiality Legal statements Trademarks

Copyright 1996–2000 Nortel Networks Corporation, All Rights Reserved

Printed in the United States of America

NORTEL NETWORKS CONFIDENTIAL

The information contained herein is the property of Nortel Networks and is strictly confidential. Except as expressly authorized in writing by Nortel Networks, the holder shall keep all information contained herein confidential, shall disclose it only to its employees with a need to know, and shall protect it, in whole or in part, from disclosure and dissemination to third parties with the same degree of care it uses to protect its own confidential information, but with no less than reasonable care. Except as expressly authorized in writing by Nortel Networks, the holder is granted no rights to use the information contained herein.

Information is subject to change without notice. Nortel Networks reserves the right to make changes in design or components as progress in engineering and manufacturing may warrant.

* Nortel Networks, the Nortel Networks logo, the Globemark HOW the WORLD SHARES IDEAS, and Unified Networks are trademarks of Nortel Networks. Trademarks are acknowledged with an asterisk (*) at their first appearance in the document.

Nortel Networks Confidential v

CDMA Technology Overview Student Guide NBSS9.0

Publication historyOctober 2000

Course updated for NBSS9.0 software release, Standard Issue 04.02. The course was modified to reflect curriculum changes, and course content was deleted and realigned:

• course content was rearranged in Lessons 1 and 3.

August 2000

Course updated for NBSS9.0 software release, Standard Issue 04.01. The course was modified to reflect curriculum changes, and course content was deleted and realigned:

• course time was changed from two days to three days

• some course content was simplified to enhance greater understanding

• List of Terms was updated

December 1999

Course updated for NBSS8.2 software release, Preliminary Issue 03.05. Course modified to reflect curriculum changes. Content deleted and realigned:

• course time reduced from five days to two days

• all Nortel Networks product information moved to course 809B

• course content aligned with new 1201 (old 810)—overlapping and advanced information was removed

September 1999

Course updated for NBSS8.0 software release, Final Issue 03.00.

January 1999

Final Issue 02.04.

vi Publication history Nortel Networks Confidential

Course 809A Standard 04.02 October 2000 For training purposes only

October 1998

Final Issue 02.03.

June 1998

Final Issue 02.02.

January 1998

Final Issue 02.01.

March 1997

Final Issue 02.00.

October 1996

Final Issue 01.00.

Nortel Networks Confidential vii


Contents 1

About this course ixCourse objectives xiPrerequisites xiCourse agenda xiSupport material xi

Lesson 1CDMA basics 1-1Objectives 1-1Terms 1-2Introduction 1-4Major subsystems of a CDMA system 1-7CDMA channels 1-9Spread spectrum principles 1-10

Lesson 2Spectrum usage and system capacity 2-1Objectives 2-1Spectrum usage and system capacity 2-2Signal strength (Eb/N0) and S/N 2-3Overlaying CDMA on an AMPS system 2-6800 MHz cellular spectrum utilization 2-7Deploying CDMA on the 1900 MHz band 2-8

Overlaying CDMA on the 1900 MHz band 2-11CDMA frequency channel assignment at 800 MHz (Cellular) 2-13What happens during a call? 2-17

Lesson 3CDMA forward channels 3-1Objectives 3-1Forward channel coding process 3-2

Sampling 3-5Quantizing 3-5

Simplified vocoder functions 3-6Signal regeneration 3-9

Pilot channel 3-26Sync channel 3-31

viii Contents Nortel Networks Confidential


Paging channels 3-41

Lesson 4CDMA reverse channels 4-1Objectives 4-1Access channels 4-4Reverse traffic channels 4-12

Lesson 5Power control, registration, and handoffs 5-1Objectives 5-1CDMA power control 5-2Registration 5-11Handoffs 5-24

Intra-system CDMA-to-analog handoff 5-36Enhanced hard handoff triggers 5-37

FiguresFigure 1-1 Why CDMA? 1-4Figure 1-2 What is Multiple Access? 1-5Figure 1-3 Multiple access technologies 1-6Figure 1-4 CDMA system components 1-7Figure 1-5 Defining our terms 1-9Figure 1-6 CDMA is a spread-spectrum system 1-10Figure 1-7 Spread spectrum principles - the “math hammer” 1-11Figure 1-8 Spread spectrum principles - many code channels 1-12Figure 1-9 CDMA spreading principle: “Anything We Can Do, We Can Undo” 1-

13Figure 1-10 “Shipping and Receiving” via CDMA 1-14Figure 1-11 CDMA’s nested spreading sequences! 1-15Figure 1-12 Walsh codes 1-16Figure 1-13 Example of correlation between Walsh Codes 1-17Figure 1-14 Correlation and Orthogonality 1-17Figure 1-15 Short PN Sequences 1-19Figure 1-16 Long PN Sequences 1-20Figure 1-17 Discriminating among forward code channels 1-21Figure 1-18 Discriminating among Base Stations 1-22Figure 1-19 Discriminating among reverse code channels 1-23Figure 1-20 Summary of characteristics and functions 1-24Figure 2-1 Spectrum usage and system capacity: Signal Bandwidth, Vulnerability,

and Frequency Reuse 2-2Figure 2-2 Relationship between Eb/N0 and S/N 2-3Figure 2-3 S/N advantage of CDMA 2-4Figure 2-4 Overlaying CDMA on an AMPS system 2-5Figure 2-5 800 MHz cellular spectrum utilization 2-7Figure 2-6 Deploying CDMA on the 1900 MHz band 2-8Figure 2-7 1900 MHz PCS spectrum utilization 2-9Figure 2-8 Overlaying CDMA on the 1900 MHz band 2-11Figure 2-9 Number of voice channels (as AMPS channels are converted to

digital) 2-12

Nortel Networks Confidential Contents ix


Figure 2-10 CDMA frequency channel assignment at 800 MHz (Cellular) 2-13Figure 2-11 CDMA frequency clearing: A-band (N=7 reuse pattern) 2-14Figure 2-12 CDMA frequency clearing: B-band (N=7 reuse pattern) 2-14Figure 2-13 Overlay guard zone deployment 2-15Figure 2-14 Other technologies: avoiding interference 2-15Figure 2-15 CDMA: using a new dimension 2-16Figure 2-16 The network view 2-17Figure 2-17 The handset view 2-17Figure 3-1 CDMA forward traffic channels 3-2Figure 3-2 CDMA code channels in the forward direction 3-3Figure 3-3 Coding process on CDMA forward code channels 3-4Figure 3-4 Digital Stream 0 (DS0) 3-5Figure 3-5 Variable rate vocoding & multiplexing (Traffic Channels only) 3-6Figure 3-6 Converting bits into symbols 3-7Figure 3-7 Spreading symbols into chips 3-8Figure 3-8 Reversing the process 3-9Figure 3-9 Forward traffic channels: Vocoding 3-10Figure 3-10 Variable rate vocoder 3-11Figure 3-11 Wireless data service 3-12Figure 3-12 Forward traffic channel generation 3-13Figure 3-13 Forward traffic channel frame structure 3-14Figure 3-14 Convolutional encoding and symbol repetition 3-15Figure 3-15 A very simple convolutional encoder 3-16Figure 3-16 Rate 1/2, K=9 convolutional encoding 3-17Figure 3-17 Symbol repetition and power reduction 3-18Figure 3-18 Symbol puncturing – Rate Set 2 (13 kb vocoder) 3-19Figure 3-19 Block interleaving 3-20Figure 3-20 9600 bps block interleaver (input array) 3-21Figure 3-21 9600 bps block interleaver (output array) 3-22Figure 3-22 9600 bps de-interleaving 3-23Figure 3-23 Forward channel demodulation 3-24Figure 3-24 Putting it all together: CDMA code channels 3-25Figure 3-25 Pilot channel 3-26Figure 3-26 Pilot channel generation 3-27Figure 3-27 Walsh Codes generation 3-28Figure 3-28 CDMA “Short” and “Long” PN codes 3-29Figure 3-29 Pilot channel acquisition 3-30Figure 3-30 Sync channel 3-31Figure 3-31 Frames and messages 3-32Figure 3-32 Sync channel generation 3-33Figure 3-33 Sync channel block interleaver (input matrix) 3-34Figure 3-34 Sync channel block interleaver (output matrix) 3-35Figure 3-35 Sync channel block interleaver (block restored) 3-36Figure 3-36 Sync channel structure 3-37Figure 3-37 Sync channel message body format 3-38Figure 3-38 Sync channel message parameters 3-39Figure 3-39 Sync channel message parameters 3-40Figure 3-40 Paging channels 3-41Figure 3-41 Paging channel generation 3-42Figure 3-42 Paging channel structure 3-43Figure 4-1 CDMA code channels in the reverse direction 4-2

x Contents Nortel Networks Confidential


Figure 4-2 Coding process on CDMA reverse code channels 4-3Figure 4-3 Access channels 4-4Figure 4-4 Access channel generation 4-5Figure 4-5 Rate 1/3 convolutional encoder 4-6Figure 4-6 Access channel block interleaving 4-7Figure 4-7 Access channel block interleaving (4800 X 2 bps – WRITE

MATRIX) 4-8Figure 4-8 Access channel block interleaving (4800 X 2 bps – READ MATRIX) 4-

9Figure 4-9 Access channel slot structure 4-10Figure 4-10 Access channel probing 4-11Figure 4-11 CDMA reverse traffic channels 4-12Figure 4-12 Reverse traffic channel generation 4-13Figure 4-13 Reverse traffic channel frame structure 4-14Figure 4-14 Reverse traffic channel: Convolutional encoding and symbol

repetition 4-15Figure 4-15 Reverse traffic channel: Block interleaving 4-16Figure 5-1 CDMA power control 5-3Figure 5-2 Reverse open loop power control 5-4Figure 5-3 Estimated reverse open loop output power 5-5Figure 5-4 Reverse closed loop power control 5-6Figure 5-5 Power output estimations (summary) 5-7Figure 5-6 Reverse outer loop power control 5-8Figure 5-7 Forward traffic channel power control 5-9Figure 5-8 Summary of all power control mechanisms 5-10Figure 5-9 Roaming 5-11Figure 5-10 HLR and VLR 5-12Figure 5-11 CDMA registration 5-13Figure 5-12 Forms of CDMA registration 5-14Figure 5-13 Power-up registration 5-15Figure 5-14 Power-down registration 5-16Figure 5-15 Timer-based registration 5-17Figure 5-16 Distance-based registration 5-18Figure 5-17 Zone-based registration 5-19Figure 5-18 Parameter-change registration 5-20Figure 5-19 Implicit registration 5-21Figure 5-20 Ordered and traffic channel registration 5-22Figure 5-21 What is Ec/Io 5-24Figure 5-22 What’s in a handset? 5-25Figure 5-23 CDMA handoffs 5-25Figure 5-24 CDMA soft handoff mechanics 5-26Figure 5-25 Softer handoff 5-27Figure 5-26 Overall handoff perspective 5-28Figure 5-27 CDMA-to-CDMA hard handoff 5-29Figure 5-28 Pilot detection trigger - CELL_PILOT_BEACON sectors 5-31Figure 5-29 Hard handoff using Beacon Pilot sectors 5-32Figure 5-30 Boundary sector trigger – border cells 5-33Figure 5-31 Hard handoff using border sectors 5-35Figure 5-32 CDMA-to-analog handoff 5-36

Nortel Networks Confidential xi


About this courseThis course presents the fundamental concepts of code division multiplex access (CDMA) theory.

Course objectives 0Upon completion of this course, the student will have a basic understanding of CDMA technology. This includes the following concepts:

• How CDMA is compared to other access technologies

• The purpose of CDMA coding, forward, and reverse channels

• Why vocoding, multiplexing, and power control is neccessary in a CDMA system

• The primary components that comprise a CDMA system

• The basic workings of CDMA messaging and call flow

Prerequisites 0This course has no prerequisites.

Course agenda 0Course material is presented in the following order:

• Lesson 1–Course basics

• Lesson 2–Spectrum usage and system capacity

• Lesson 3–CDMA forward channels

• Lesson 4–CDMA reverse channels

• Lesson 5–Power control, registration, and handoffs

Support material 0Each student should have a copy of the CDMA Technology Overview Student Guide.

xii About this course Nortel Networks Confidential


Nortel Networks Confidential 1-1

CDMA Technology Overview NBSS9.0

Lesson 1 CDMA basicsObjectives 1

Upon completion of this lesson, the student will be able to:

• compare CDMA to other access technologies

• describe the concept of spectrum spreading

• explain how CDMA uses Walsh Codes

• distinguish between short and long PN codes

• explain the relationship between bits, symbols, and chips

1-2 CDMA basics Nortel Networks Confidential


Terms 1The following terms are used in this lesson:

AMPS—advanced mobile phone service

AM—amplitude modulation

BER—bit error rate

BTA—basic trading area

CDMA—code division multiplex access

CRC—cyclic redundancy check

CTIA—Cellular Telecommunications Industry Association

D-AMPS—digital American phone system

DSP—digital signal processor

ESN—electronic serial number

FDMA—frequency division multiplex access

FER—frame error rate

FSK—frequency shift keying

GSM—global system for mobile communication

kb—kilobit

LAN—local area network

LEC—local exchange carrier

MHz—Megahertz

MTA—multiple trading area

PCS—personal communication service

PN—pseudo-random number

PSTN—public switched telephone network

Nortel Networks Confidential CDMA basics 1-3


QCELP—Qualcomm code excited linear prediction

QPSK—quadrature phase shift keying

RF—radio frequency

VSELP—vector sum excited linear prediction

Walsh Codes—Forward CDMA channels received at a mobile from a sector in a base station

XOR—exclusive OR



Introduction 1Code division multiplex access (CDMA) is a spread spectrum system developed by Qualcomm. The technology has gained wide acceptance for both cellular and PCS, largely due to the fact that it is extremely robust with excellent audio quality from 8 kb and 13 kb variable rate vocoders.

Characteristics of CDMA include the following:

• Is the technology of choice for both 800 MHz Cellular and 1900 MHz PCS service providers

• Satisfies CTIA users’ performance requirements

• Provides high capacity (in excess of 10 x AMPS)

• Provides excellent audio quality

• Provides privacy through its coding scheme

Figure 1-1Why CDMA?

■ Is the technology of choice forboth 800 MHz Cellular and1900 MHz PCS serviceproviders

■ Satisfies CTIA Users’Performance Requirements

■ Provides high capacity (manytimes the capacity of AMPS)

■ Provides privacy through itscoding scheme

CDMA

CDMA

ode

ivision

ultipleccess

CDMA is extremely robust andprovides excellent audio quality



Figure 1-2What is Multiple Access?

Multiple access techniques are used to increase efficiency in a variety of settings. Commonplace examples that can be encountered directly or indirectly today include the following:

• local area networks (LAN)

Computers in medium and large businesses are usually linked together over a shared media (usually twisted pair cable). This facilitates multiple access to applications and data, and the ability to exchange files easily. Typical LAN configurations are Token Ring or Ethernet.

• local telephone service

The typical home subscriber accesses the public switched telephone network (PSTN) over time division multiplexed lines. These multiplexed lines concentrate individual lines and feed them to the local office. Increasingly today, the physical media of choice is fiber optic cable. The use of fiber decreases operational costs for the local exchange carrier (LEC) and broadens the range of services that can be supported.

CDMA Technology Overview February, 2000 - Page 1-4

What is Multiple Access?

Since the beginning of telephony and radio, system operators have tried to squeeze the maximum amount of traffic over each circuitTypes of Media

• Twisted pair - copper• Coaxial cable• Fiber optic cable• Air interface (radio signals)

Advantages of Multiple Access• Increased capacity: serve more users• Reduced capital requirements since fewer

media can carry the traffic• Decreased per-user expense• Easier to manage and administer

Transmission

Medium

Each pair of users enjoys a dedicated, private circuit through the transmission medium, unaware that the

other users exist.

Multiple Access: Simultaneous private use of a transmission medium by multiple, independent users.



Figure 1-3Multiple access technologies

Frequency division multiplex access (FDMA) is the oldest and most widely implemented access method. It makes the least efficient use of the limited frequency spectrum, allocating each user a fixed band of frequencies for their exclusive use for the duration of a call.

Time division multiplex access (TDMA) is used by D-AMPS and GSM, and makes better use of the same spectrum than AMPS by allowing several users to share the same band of frequencies. Instead of assigning a frequency bands to just one user, TDMA divides the time into slots and shares the channel between several users by assigning them different slots in a round-robin basis. This method is still inefficient (although less than AMPS) because the time slots allocated to one conversation cannot be re-allocated for another purpose if they are not used by their original designated user.

Code division multiplex access (CDMA) is a spread spectrum modulation technology where channels are defined by means of mathematical codes, and which share the same frequency band simultaneously.


Multiple Access Technologies

The physical transmission medium is a resource that can be subdivided into individual channels according to different criteria depending on the technology used:

Here’s how the three most popular technologies establish channels:

• FDMA (Frequency Division Multiplex Access)− each user on a different frequency− a channel is a frequency

• TDMA (Time Division Multiplex Access)− each user on a different window period in time

(“time slot”)− a channel is a specific time slot on a specific

frequency• CDMA (Code Division Multiplex Access)

− each user uses the same frequency all the time, but mixed with different distinguishing code patterns

− a channel is a unique set of code patterns FrequencyTime

Power

FrequencyTime

Power

FrequencyTime

Power

FDMA

TDMA

CDMA

Channel: An individually-assigned, dedicated pathway through a transmission medium for one user’s information



Figure 1-4CDMA system components

Major subsystems of a CDMA system 1The hardware architecture of the CDMA wireless system consist of switching equipment and cell site equipment. These components interact with the Public Switched Telephone Network (PSTN) and the Mobile Subscriber Unit (MSU) to provide a complete cellular communications system.

The major subsystems of the Nortel Networks product design and function are:

• Digital Multiplex System-Mobile Telephone Exchange (DMS-MTX) provides the high level call processing for multiple access technoloies (AMPS, TDMA, CDPD and CDMA). The DMS-MTX will make and break many thousands of connections during normal operations, and performs the following functions:

— records billing information

— performs diagnostics

— generates logs

— records call information

• Mobile Telephone Exchange (MTX) provides call processing functions for AMPS/TDMA/CDPD/CDMA cellular systems

• Base Station Subsystem Manager (BSSM) provides a Graphical User Interface (GUI) for operations, administration and maintenance of the BSC, BTS and itself

• Base Station Controller (BSC) provides data routing, voice coding and some hand-off functions

• Base Station Transceiver Subsystem (BTS) provides the RF link to the subscriber

• MTX, BSC and BSM are identical for 800 and 1900 MHz products

• Mobile Telephone Exchange (MTX) provides call processing functions for AMPS/TDMA/CDPD/CDMA cellular systems

• Base Station Subsystem Manager (BSSM) provides a Graphical User Interface (GUI) for operations, administration and maintenance of the BSC, BTS and itself

• Base Station Controller (BSC) provides data routing, voice coding and some hand-off functions

• Base Station Transceiver Subsystem (BTS) provides the RF link to the subscriber

• MTX, BSC and BSM are identical for 800 and 1900 MHz products

DMS-MTX BTS

T1 or E1s

MTSO

T1s

BSSM BSCMAP



— records statistics

— routes calls

— validates users

• Base Station Subsystem Manager (BSSM) provides a Graphical User Interface (GUI) for the operations, administration and maintenance of the Base Station Controller (BSC) and Base Station Transceiver Subsystem (BTS). The BSSM performs the following functions:

— software download

— initialize and enable network elements

— software and configuration file storage

— database collection and analysis

— monitoring, testing and diagnosis

— performance analysis

— system administration

• Base Station Controller (BSC) controls the message and signalling routing between itself, the DMS-MTX, BSSM and BTS. It also provides the voice coding and decoding between the MSU via the BTS and the PCM from the DMS-MTX. Only one BSC is required for the system and it is normally co-located with the DMS-MTX. The BSC provides call processing functions such as:

— power control

— service options

— soft handoffs

• Base Station Transceiver Subsystem (BTS) provides the over-the-air interface to the Mobile Subscriber Unit (MSU), according to the IS-95A for 800MHz (Indoor Product) and ANSI J-STD-008 for the 1900 MHz (Outdoor Product) with the DMS-MTX. The BTS equipment location, frequently referred to as a cell site, can contain multiple BTSs operating at different frequency assignments. The BTS can be configured as a Omni, 2-Sector or a 3-Sector cell, and performs the following functions:

— converts digital vocoded HDLC Packet into RF signals

— converts RF signals into digital vocoded HDLC Packet

— power control

— softer handoffs



Figure 1-5Defining our terms

CDMA channels 1CDMA subdivides the available spectrum in 1.25 MHz bands referred to as CDMA Channels or CDMA carriers. Each of these carriers can typically support 22 or 17 calls (depending on whether 8 or 13 kb vocoders are used), per sector; in addition to supporting handoff with about one third to one half as many mobile stations nominally being serviced by other cells.

In addition to the forward and reverse traffic channels, other overhead or support channels (pilot, sync, paging, and access) coexist in the same 1.25 MHz carrier. Multiple carriers can be accommodated in a licensed block of spectrum.

The individual channels that share the same CDMA channel, or carrier, are referred to as forward, or reverse code channels. Each code channel behaves as a source of noise from the point of view of the other, and the maximum number of such channels that can coexist on a carrier at any given time is not fixed, but depends on the total level of noise against which each code channel has to contend. It is for this reason that a CDMA system is said to have a soft capacity limit.


Defining Our Terms

■ CDMA Channel or CDMA Carrier or CDMA Frequency• Duplex channel made of two 1.25 MHz-wide bands of electromagnetic

spectrum, one for Base Station to Mobile Station communication (called the FORWARD LINK or the DOWNLINK) and another for Mobile Station to Base Station communication (called the REVERSE LINK or the UPLINK)

• In 800 MHz Cellular these two simplex 1.25 MHz bands are 45 MHz apart• In 1900 MHz PCS they are 80 MHz apart

■ CDMA Forward Channel

• 1.25 MHz Forward Link

■ CDMA Reverse Channel• 1.25 MHz Reverse Link

■ CDMA Code Channel• Each individual stream of 0’s and 1’s contained in either the CDMA

Forward Channel or in the CDMA Reverse Channel• Code Channels are characterized (made unique) by mathematical codes• Code channels in the forward link: Pilot, Sync, Paging and Forward Traffic

channels• Code channels in the reverse link: Access and Reverse Traffic channels

45 or 80 MHz

CDMA CHANNELCDMA

ReverseChannel 1.25 MHz

CDMAForwardChannel 1.25 MHz



Figure 1-6CDMA is a spread-spectrum system

Spread spectrum principles 1Traditional radio communication systems transmit data using the minimum bandwidth required to carry it as a narrowband signal. Direct-Sequence Spread Spectrum systems mix their input data with a fast spreading sequence and transmit a wideband signal.

The spreading sequence is independently regenerated at the receiver and mixed with the incoming wideband signal to recover the original data. The de-spreading gives substantial gain proportional to the bandwidth of the spread-spectrum signal.

The gain can be used to increase system performance and range, or allow multiple coded users, or both.

A digital bit stream sent over a radio link requires a definite bandwidth to be successfully transmitted and received. The actual spectrum occupied by a digital signal depends on its bit rate and type of radio modulation (AM, FSK, QPSK, etc.).


CDMA Is a Spread-Spectrum System

• Traditional technologies try tosqueeze the signal into the minimumrequired bandwidth

• Direct-Sequence Spread spectrumsystems mix their input data with afast spreading sequence andtransmit a wideband signal

• The spreading sequence isindependently regenerated at thereceiver and mixed with theincoming wideband signal to recoverthe original data

• The de-spreading gives substantialgain proportional to the bandwidth ofthe spreading signal

• CDMA uses a larger bandwidth butthen uses resulting processing gainto increase capacity

Spread Spectrum Payoff:Processing Gain

Spread SpectrumTRADITIONAL COMMUNICATIONS SYSTEM

SlowInformation

Sent

TX

SlowInformationRecovered

RX

NarrowbandSignal

SPREAD-SPECTRUM SYSTEM

FastSpreadingSequence

SlowInformation

Sent

TX

SlowInformationRecovered

RX

FastSpreadingSequence

Wideband Signal



• The minimum usable bandwidth (in Hertz) is roughly equal to the speed of the information (in bits per second).

• Only very advanced modulation forms achieve this density.

• Common techniques use several times as much radio bandwidth as information bandwidth.

Most of radio history has been a series of attempts to squeeze more and more information into smaller and smaller channel bandwidths, thus making room for more users. However, CDMA exploits a different principle:

If a signal is transmitted deliberately using much more RF bandwidth than really required, it will be much easier to detect at the receiver.

There is actually a reward for wasting spectrum! This reward is called Processing Gain. CDMA uses it to make the RF link more reliable, and uses coding to allow multiple users to share the same very wideband signal. This deliberate technique is called Spread Spectrum transmission.

Figure 1-7Spread spectrum principles - the “math hammer”

1.25 MHz30 KHz

Power is “Spread” Over a Larger BandwidthMATHHAMMER

MATHHAMMER



Figure 1-8Spread spectrum principles - many code channels

Many code channels are individually“spread” and then added together tocreate a “composite signal”

UNWANTED POWERFROM OTHER SOURCES

Using the “right” mathematicalsequences any Code Channelcan be extracted from the receivedcomposite signal



Figure 1-9CDMA spreading principle: “Anything We Can Do, We Can Undo”


Anything We Can Do, We Can Undo

■ Any data bit stream can be combined with a spreading sequence

■ The resulting signal can be de-spread and the data streamrecovered if the original spreading sequence is available andproperly synchronized

■ After de-spreading, the original data stream is recovered intact

ORIGINATING SITE DESTINATION

SpreadingSequence

SpreadingSequence

InputData

(Base Band)

RecoveredData

(Base Band)

Spread Data Stream(Base Band + Spreading Sequence)



Figure 1-10“Shipping and Receiving” via CDMA

■ Whether in shipping and receiving, or in CDMA,packaging is extremely important!

■ Cargo is placed inside “nested” containers for protectionand to allow addressing

■ The shipper packs in a certain order, and the receiverunpacks in the reverse order

■ CDMA “containers” are spreading codes

Fed

Ex

Data Mailer

Fed

Ex

DataMailer

Shipping Receiving



Figure 1-11CDMA’s nested spreading sequences!

■ CDMA combines three different spreading sequences to createunique, robust channels

■ The sequences are easy to generate on both sending andreceiving ends of each link

■ The sequences are applied in succession at the sending end andthen reapplied in opposite order to recover the original data streamat the receiving end

SpreadingSequence

A

SpreadingSequence

B

SpreadingSequence

C

SpreadingSequence

C

SpreadingSequence

B

SpreadingSequence

A

InputData

X

RecoveredData

X

X+A X+A+B X+A+B+C X+A+B X+ASpread-Spectrum Chip Streams

ORIGINATING SITE DESTINATION



Figure 1-12Walsh codes

WALSH CODES # ---------------------------------- 64-Chip Sequence ------------------------------------------ 0 0000000000000000000000000000000000000000000000000000000000000000 1 0101010101010101010101010101010101010101010101010101010101010101 2 0011001100110011001100110011001100110011001100110011001100110011 3 0110011001100110011001100110011001100110011001100110011001100110 4 0000111100001111000011110000111100001111000011110000111100001111 5 0101101001011010010110100101101001011010010110100101101001011010 6 0011110000111100001111000011110000111100001111000011110000111100 7 0110100101101001011010010110100101101001011010010110100101101001 8 0000000011111111000000001111111100000000111111110000000011111111 9 010101011010101001010101101010100101010110101010010101011010101010 001100111100110000110011110011000011001111001100001100111100110011 011001101001100101100110100110010110011010011001011001101001100112 000011111111000000001111111100000000111111110000000011111111000013 010110101010010101011010101001010101101010100101010110101010010114 001111001100001100111100110000110011110011000011001111001100001115 011010011001011001101001100101100110100110010110011010011001011016 000000000000000011111111111111110000000000000000111111111111111117 010101010101010110101010101010100101010101010101101010101010101018 001100110011001111001100110011000011001100110011110011001100110019 011001100110011010011001100110010110011001100110100110011001100120 000011110000111111110000111100000000111100001111111100001111000021 010110100101101010100101101001010101101001011010101001011010010122 001111000011110011000011110000110011110000111100110000111100001123 011010010110100110010110100101100110100101101001100101101001011024 000000001111111111111111000000000000000011111111111111110000000025 010101011010101010101010010101010101010110101010101010100101010126 001100111100110011001100001100110011001111001100110011000011001127 011001101001100110011001011001100110011010011001100110010110011028 000011111111000011110000000011110000111111110000111100000000111129 010110101010010110100101010110100101101010100101101001010101101030 001111001100001111000011001111000011110011000011110000110011110031 011010011001011010010110011010010110100110010110100101100110100132 000000000000000000000000000000001111111111111111111111111111111133 010101010101010101010101010101011010101010101010101010101010101034 001100110011001100110011001100111100110011001100110011001100110035 011001100110011001100110011001101001100110011001100110011001100136 000011110000111100001111000011111111000011110000111100001111000037 010110100101101001011010010110101010010110100101101001011010010138 001111000011110000111100001111001100001111000011110000111100001139 011010010110100101101001011010011001011010010110100101101001011040 000000001111111100000000111111111111111100000000111111110000000041 010101011010101001010101101010101010101001010101101010100101010142 001100111100110000110011110011001100110000110011110011000011001143 011001101001100101100110100110011001100101100110100110010110011044 000011111111000000001111111100001111000000001111111100000000111145 010110101010010101011010101001011010010101011010101001010101101046 001111001100001100111100110000111100001100111100110000110011110047 011010011001011001101001100101101001011001101001100101100110100148 000000000000000011111111111111111111111111111111000000000000000049 010101010101010110101010101010101010101010101010010101010101010150 001100110011001111001100110011001100110011001100001100110011001151 011001100110011010011001100110011001100110011001011001100110011052 000011110000111111110000111100001111000011110000000011110000111153 010110100101101010100101101001011010010110100101010110100101101054 001111000011110011000011110000111100001111000011001111000011110055 011010010110100110010110100101101001011010010110011010010110100156 000000001111111111111111000000001111111100000000000000001111111157 010101011010101010101010010101011010101001010101010101011010101058 001100111100110011001100001100111100110000110011001100111100110059 011001101001100110011001011001101001100101100110011001101001100160 000011111111000011110000000011111111000000001111000011111111000061 010110101010010110100101010110101010010101011010010110101010010162 001111001100001111000011001111001100001100111100001111001100001163 0110100110010110100101100110100110010110011010010110100110010110



Figure 1-13Example of correlation between Walsh Codes

There are 64 Walsh Codes used in the CDMA system. Each sequence is 64 chips in length, where a chip is a binary digit (1 or 0).

Each Walsh Code is orthogonal to the other codes: if the result of XORing two Walsh codes results in the same number of 0s and 1s, they are said to be orthographic. If these binary strings were represented as lines, they would be perpendicular to each other.

Figure 1-14Correlation and Orthogonality

EXAMPLE:

Correlation of Walsh Code #23 with Walsh Code #59

#23 0110100101101001100101101001011001101001011010011001011010010110#59 0110011010011001100110010110011010011001011001100110011010011001XOR 0000111111110000000011111111000011110000000011111111000000001111

Correlation Results: 32 1’s, 32 0’s: Orthogonal!!

CDMA Theory and Product Design September, 1999 - Page 1-16

Correlation and OrthogonalityCorrelation and Orthogonality

Correlation is a measure of the similarity between two binary strings

Code #23 0110100101101001100101101001011001101001011010011001011010010110

–(Code #23) 1001011010010110011010010110100110010110100101100110100101101001

Code #59 0110011010011001100110010110011010011001011001100110011010011001

PARALLELPARALLEL

XOR: all 0sXOR: all 0s

Correlation: 100%(100% match)

Correlation: 100%(100% match)

ORTHOGONALORTHOGONAL

XOR: half 0s, half 1sXOR: half 0s, half 1s

Correlation: 0%(50% match, 50% no-match)

Correlation: 0%(50% match, 50% no-match)

ANTI-PARALLELANTI-PARALLEL

XOR: all 1sXOR: all 1s

Correlation: –100%(100% no-match)

Correlation: –100%(100% no-match)

#23#23

–(#23)

#23

#23

#59



The inherent orthographic structure of the Walsh Code table makes it easier to recognize and extract a particular code when multiple codes are being transmitting simultaneously. The codes that are not orthogonal to the code being analyzed are discarded.



Figure 1-15Short PN Sequences


The Short PN SequencesThe Short PN Sequences

The two Short PN Sequences, I and Q,are 32,768 chips long

• Together they can be considered atwo-dimensional binary “vector” withdistinct I and Q component sequences,each 32,768 chips long

• Each Short PN Sequence (and, as amatter of fact, any sequence)correlates with itself perfectly ifcompared at a timing offset of 0 chips

• Each Short PN Sequence is special:Orthogonal to a copy of itself that hasbeen offset by any number of chips(other than 0)

IQ

32,768 chips long26 2/3 ms.

(75 repetitions in 2 sec.)

IQIQ

100% Correlation: All bits = 0

Short PN Sequence vs. Itself @ 0 Offset

IQIQ

Orthogonal: 16,384 1’s + 16,384 0’s

Short PN Sequence vs. Itself @ Any Offset

Unique Properties:



Figure 1-16Long PN Sequences


The Long PN SequenceThe Long PN Sequence

• Each mobile station uses a unique User Long Code Sequence generatedby applying a mask, based on its 32-bit ESN, to the 42-bit Long CodeGenerator which was synchronized with the CDMA system during themobile station initialization

• Generated at 1.2288 Mcps, this sequence requires 41 days, 10 hours, 12minutes and 19.4 seconds to complete

• Portions of the Users Long Codes generated by different mobile stationsfor the duration of a call are not exactly orthogonal but are sufficientlydifferent to permit reliable decoding on the reverse link

Long Code Register (@ 1.2288 MCPS)

Public Long Code Mask (STATIC)

User Long CodeSequence

(@1.2288 MCPS)

1 1 0 0 0 1 1 0 0 0 PERMUTED ESNAND

=SUM

Modulo-2 Addition



Figure 1-17Discriminating among forward code channels

■ A Mobile Station, tuned to a particular CDMA frequency, receives a ForwardCDMA Channel from a sector in a Base Station.

■ This Forward CDMA Channel carries a composite signal made of up to 64“forward code channels”

■ Some of these code channels are “traffic channels” while other are“overhead channels” needed by the CDMA system to operate properly.

■ A set of 64 mathematical codes is needed to differentiate the 64 possibleforward code channels that can be contained in a Forward CDMA Channel.

• The codes in this set are called “Walsh Codes”

SyncPilot

FW Traffic(for user #1)

Paging





Figure 1-18Discriminating among Base Stations

■ A mobile Station is surrounded by Base Stations, all of them transmittingon the same CDMA Frequency

■ Each Sector in each Base Station is transmitting a CDMA Forward TrafficChannel containing up to 64 distinct forward code channels

■ A Mobile Station must be able to discriminate between different Sectors ofdifferent Base Stations and listen to only one set of code channels

■ Two binary digit sequences called the I and Q Short PN Sequences (orShort PN Codes) are defined for the purpose of identifying sectors ofdifferent base stations

■ These Short PN Sequences can be used in 512 different ways in a CDMAsystem. Each one of them constitutes a mathematical code which can beused to identify a particular sector of a particular base station

A B

Up to 64Code Channels

Up to 64Code Channels



Figure 1-19Discriminating among reverse code channels

■ The CDMA system must be able touniquely identify each Mobile Station thatmay attempt to communicate with a BaseStation

■ A very large number of Mobile Stationswill be in the market

■ One binary digit sequence called theLong PN Sequence (or Long PN Code)is defined for the purpose of uniquelyidentifying each possible reverse codechannel

■ This sequence is extremely long and canbe used in trillions of different ways.Each one of them constitutes amathematical code which can be used toidentify a particular user (and is thencalled a User Long Code) or a particular“access channel” (explained later in thiscourse)

RV Trafficfrom M.S.

#1837732008RV Trafficfrom M.S.

#8764349209

RV Trafficfrom M.S.

#223663748

System AccessAttempt by M.S.#4348769902

(on access channel #1)



Figure 1-20Summary of characteristics and functions


Summary of Characteristics & FunctionsSummary of Characteristics & Functions

Cell

• Each CDMA spreading sequence is used for a specific purpose on the forward link and a different purpose on the reverse link

• The sequences are used to form “code channels” for users in both directions

Walsh Codes

Short PN Sequences

Long PNSequences

Type of Sequence

Mutually Orthogonal

Orthogonal with itself at any time shift value except 0

near-orthogonal if shifted

Special Properties

64

2

1

How Many

64 chips1/19,200

sec.

32,768 chips

26-2/3 ms75x in 2

sec.

242 chips~41 days

Length

Orthogonal Modulation(information

carrier)

Quadrature Spreading

(Zero offset)

Distinguish users

Reverse Link

Function

User identity

within cell’s signal

Distinguish Cells & Sectors

Data Scrambling to avoid all 1’s or 0’s

Forward Link

Function

IQ

32,768 chips long26-2/3 ms.

(75 repetitions in 2 sec.)

64codes

64 chips long

AND

=SUM

Modulo-2 Addition



Exercise 1-1 Lesson review

Answer the following questions and review your answers with the instructor.

1. If a signal is deliberately transmitted using more RF bandwidth than required, it is easier to detect at the receiver. This “waste” is formally defined as what?

2. Are all CDMA Walsh Codes orthogonal?

3. What sequence best describes this conversion relationship in CDMA:

a. symbols ⇒ bits ⇒ chips

b. chips ⇐ symbols ⇐ bits

c. Bits ⇒ chips ⇒ symbols

4. List the four overhead (support) channels.

5. What is the size (in MHz) of a CDMA channel?

6. What is the size (in MHz) of a CDMA guard band?





Lesson 2 Spectrum usage and system capacityObjectives 2


• understand the capacity issues related to CDMA, TDMA, and AMPS

• explain what a system overlay is

• understand the call events of what happens during a wireless call

• explain what EbNo is

2-2 Spectrum usage and system capacity Nortel Networks Confidential


Figure 2-1Spectrum usage and system capacity: Signal Bandwidth, Vulnerability, and Frequency Reuse

Spectrum usage and system capacity 2All wireless technologies depend on frequency reuse to multiply capacity. If the radio signal is fragile, a high carrier to interference ratio (C/I) is required and cells sharing the same frequency must be physically far apart. The same frequency cannot be used in adjacent cells, other cells are required to fill in the gaps, and each cell uses only a fraction of the total available frequencies. If the radio signal is robust, or uses special techniques to distinguish users, then cells sharing the same frequencies can be closer: each cell can have more frequencies and therefore more users.

To reduce interference in AMPS and TDMA, in order to achieve the required level of C/I, we must increase the distance between co-channel cells D relative to their coverage radio R. This imposes a reuse factor N which results in a proportional decrease of cell capacity (where capacity expressed as the number of channels per cell is calculated as the total number of channels divided by the reuse factor N).

In AMPS and North American TDMA, a reuse factor of N = 7 is the minimum that “almost” works in most propagation environments, giving about 17 dB of C/I and 56 channels per cell (assuming a total of 395 traffic

■ Each wireless technology (AMPS,NAMPS, D-AMPS, GSM, CDMA) uses aspecific modulation type with its ownunique signal characteristics

■ The total traffic capacity of a wirelesssystem is determined largely by radiosignal characteristics and RF design

■ RF signal vulnerability to Interferencedictates how much interference can betolerated, and therefore how far apartsame-frequency cells must be spaced

■ For a specific S/N level, the SignalBandwidth determines how many RFsignals will “fit” in the operator’s licensedspectrum

AMPS, D-AMPS, N-AMPS

CDMA

30 30 10 kHz

200 kHz

1250 kHz

1 3 1 Users

8 Users

22 Users1

1

11

1

11

11

1

11

1

1

12

34

43

2

56

17

Typical Frequency Reuse N=7



Vulnerability:C/I ≅ 17 dB

Vulnerability:C/I ≅ 12-14 dB

Vulnerability:Eb/No ≅ 6 dB

GSM

17 dB = 101.7 ≅ 5014 dB = 101.4 ≅ 25 12 dB = 101.2 ≅ 16

Nortel Networks Confidential Spectrum usage and system capacity 2-3


channels). The C/I can be further improved by subdividing the cell in three sectors and giving each sector one third of the channels in that cell, but this comes at a price: the reduction of traffic capacity measured in Erlangs.

As GSM can operate with a lower C/I, a frequency reuse factor of four can be successfully used.

This problem does not exist in CDMA where a reuse factor of one is the norm.

Figure 2-2Relationship between Eb/N0 and S/N

Signal strength (Eb/N0) and S/N 2The average amount of Energy per Bit of information (Eb) can be calculated dividing the total signal power (energy per unit of time) by the bit rate (number of bits transmitted per unit of time).

The average Noise Spectral Density (N0) can be calculated dividing the total noise power by the bandwidth of the channel.

Eb =S

R

Signal Power

Bit Rate = N0 =

N

W

Noise Power

Bandwidth=

=S

R

W

N X =

S

N

W

R X

S

R

N

W

Eb

N0

=

Signal to Noise

ProcessingGain

E / t

B / t=

WR

=1,250,00014,400

= 87 =1.9410 = 19.4dB

WR

=1,250,000

9,600= 130 =

2.1110 = 21.1dB8 Kb vocoder(Full Rate)

13 Kb vocoder(Full Rate)



Simple math can be used combining both definitions to obtain a relationship between the rate of Energy per Bit of information to Noise Spectral Density (Eb/N0) and the rate of Signal to Noise (S/N).

The rate of Bandwidth to Bit Rate (W/R) which appears in this equation is defined as the Processing Gain. Notice that according to this definition, a CDMA system in which 9600 bps of information are transmitted using a bandwidth of 1.2288 MHz, has a processing gain of 21.1 dB, and a CDMA system in which 14400 bps of information are transmitted using a bandwidth of 1.2288 MHz, has a processing gain of 19.4dB.

Figure 2-3S/N advantage of CDMA

In CDMA, when using the 13 Kb vocoder at full rate, noise can be 22 times stronger than the signal, and the signal can still be recovered!

To make CDMA work, it is necessary not only to devise a method of making each signal sharing the bandwidth look (and behave statistically) like noise, but to do it in such a way that it is possible later to selectively recover the desired signal making it “stand above the noise.”

AMPS

N-AMPS

D-AMPS

GSM

CDMA

Analog FM

Analog FM

DQPSK

GMSK

QPSK/OQPSK

30 KHz.

10 KHz.

30 KHz.

200 KHz.

1,250 KHz.

C/I ≅ 17 dB

C/I ≅ 17 dB

C/I ≅ 17 dB

C/I ≅ 12-14 dB

Eb/No ≅ 6dB

Tech-nology Modulation Type Channel

BandwidthQuality

Indicator

S/N ≅ 17 dB

S/N ≅ 17 dB

S/N ≅ 17 dB

S/N ≅ 12 to 14 dB

S/N ≅ –13.4 dB

S/N

17 dB = 101.7 ≅ 5014 dB = 101.4 ≅ 25 12 dB = 101.2 ≅ 16

-13.4 dB = 10-1.34 ≅ 0.046 =

S

N⇒ 10 0.6

10 1.94= = 10 -1.34 = -13.4 dB

Signal to NoiseProcessing Gain (W/R)

S

NX = 10 0.610 1.94

Eb N0

122



The Signal-to-Noise rate is not important to CDMA (actually the level of a particular signal is dwarfed by the noise). Instead, what is critical for CDMA is the Eb/N0 rate; that is, the amount of energy used to represent each bit of information (carried over the CDMA channel bandwidth) relative to the noise spectral density (or amount of noise energy by unit of bandwidth).

The Bit Error Rate (BER), or number of bits in error per unit of time, is not that important either because the bit coding scheme used allows a large number of individual bit errors to be corrected automatically.

The bit stream is divided into 20 ms slices which are processed and converted into frames. It is the Frame Error Rate (FER) that is used instead of the BER as a performance metric.

Figure 2-4Overlaying CDMA on an AMPS system

Each CDMA Channel: 1.250 MHz ÷ 30 kHz = 41.7 = ~41 AMPS channelsEach Guard Band: 260 kHz ÷ 30 kHz = 8.7 = ~9 AMPS channels

260 KHz 260 KHz1.25 MHz Nominal Bandwidth

Frequency

Po

wer

1.77 MHz

CDMA CARRIER

41 AMPS channels 41 AMPS channels

9 AMPSchannels

41 AMPS channels

41 AMPS channels

CDMA CDMA CDMA CDMA CDMA

AVAILABLE AVAILABLE

885 KHzMinimum Separation between AMPS/TDMA and CDMA center frequency:

(1,250 kHz ÷ 2) + 260 kHz = 885 kHz

TOTAL: 1.77 MHz ÷ 30 kHz = 59 AMPS channels

GUARDBAND

GUARDBAND

Each CDMA Channel: 1.250 MHz ÷ 30 kHz = 41.7 = ~41 AMPS channelsEach Guard Band: 260 kHz ÷ 30 kHz = 8.7 = ~9 AMPS channels


Frequency

Po

wer

1.77 MHz


Frequency

Po

wer

1.77 MHz

CDMA CARRIER


9 AMPSchannels

41 AMPS channels

41 AMPS channels


AVAILABLE AVAILABLE


(1,250 kHz ÷ 2) + 260 kHz = 885 kHz


9 AMPSchannels

41 AMPS channels

41 AMPS channels


AVAILABLE AVAILABLE


(1,250 kHz ÷ 2) + 260 kHz = 885 kHz

TOTAL: 1.77 MHz ÷ 30 kHz = 59 AMPS channels

GUARDBAND

GUARDBAND



Overlaying CDMA on an AMPS system 2If the 1.25 MHz band of frequencies identified in the drawing as the nominal band is used to support a CDMA channel in some cells, neither the frequencies in this band nor those in the 260 kHz Guard Bands on either side of it can be used in those cells or in the cells in the surrounding Guard Zone (discussed later) for any other purpose.

Frequencies outside this 1.77 MHz band (1.25 MHz nominal plus two 260 kHz guard bands) can be used anywhere, including the central CDMA cells and the surrounding Guard Zone.

Frequencies in this 1.77 MHz band can be used in cells beyond the Guard Zone.

Notice that about 59 AMPS channels must be cleared to introduce one CDMA channel, that is, 41 for the actual CDMA signal and nine for each Guard Band.



Figure 2-5800 MHz cellular spectrum utilization

800 MHz cellular spectrum utilization 2The following table is an example of CDMA channel allocation, in chronological order, which allows maximum CDMA channel packing.

Order Side “A” Side “B”

1 283 384

2 242 425

3 201 466

4 160 507

5 119 548

6 78 589

7 37 630

8 1019 777

9 691 736

■ All CDMA RF carriers are 1.25 MHz. wide• can serve ~22 users w/8 kb vocoder (~17 users w/13 kb vocoder)

■ The cellular spectrum of one operator is 12.5 MHz. wide. You’d expectthat 10 CDMA carriers would fit. However, only 9 carriers can be used

• operators must maintain a “token” AMPS presence for several years• “guard bands” are required at the edges of frequency blocks or any

frequency boundaries between CDMA/non-CDMA signals• no guard bands are required between adjacent CDMA carriers

Possible CDMA Center Freq. Assignments

Channel Numbers

Forward link (i.e., cell site transmits)Reverse link (i.e., mobile transmits)824MHz

849MHz

869MHz

894MHz

otherusesA” A”A B A’ B’

1 10 10 1.5 2.5

A B A’ B’

1 10 10 1.5 2.5

991

10231 333

334

666667

716717

799

991

10231 333

334

666667

716717

799

~300 kHz. “guard bands” possibly required if adjacent-frequency signals are non-CDMA (AMPS, TDMA, ESMR, etc.)



Deploying CDMA on the 1900 MHz band 2

Figure 2-6Deploying CDMA on the 1900 MHz band

CDMA

1770 ÷ 50 = 35.4

1250 ÷ 50 = 25

260 ÷ 50 = 5.2

260 kHz 260 kHz1.25 MHz Nominal Bandwidth

Frequency

Po

we

r

1.77 MHz

CDMA CARRIER

• All frequenciesare available fornon-CDMA use

• Only thefrequencies in thegray area areavailable for nonCDMA use

• All frequenciesare available fornon-CDMA use

• Only thefrequencies in thegray area areavailable for nonCDMA use



Figure 2-71900 MHz PCS spectrum utilization

The following tables are examples of CDMA channel allocation, in chronological order, which allow maximum CDMA channel packing. Each table represents the “preferred” set of CDMA channels according to J-STD-008.

PCS Band A1 25

2 50

3 75

4 100

5 125

6 150

7 175

8 200

9 225

10 250

11 275

■ A, B, and C licenses can accommodate 11 CDMA RF channels in their 30 MHz of spectrum

■ D, E, and F licenses can accommodate 3 CDMA RF channels in their 10 MHz of spectrum

■ 260 kHz guard bands are required on the edges of the PCS spectrum to ensure no interference occurs with other applications just outside the spectrum

Guard Bands

Forward link (i.e., cell site transmits)Reverse link (i.e., mobile transmits)1850MHz

BTA

BTA

BTA

BTA

BTA

BTA

Paired Bands

MTA BTAMTABTA MTAMTA

1910MHz

1930MHz

1990MHz

Data Voice

A D B E F C A D B E F C

15 51010 1515151515 555 55

Licensed Licensed

Unlicensed

0

Channel Numbers 299

300

400

699700

800

900

1199 0

299300

400

699700

800

900

1199

Guard BandsGuard Bands

Forward link (i.e., cell site transmits)Reverse link (i.e., mobile transmits)1850MHz1850MHz

BTA

BTA

BTA

BTA

BTA

BTA

BTA

BTA

BTA

BTA

BTA

BTA

Paired Bands

MTA BTAMTABTA MTAMTA

1910MHz1910MHz

1930MHz1930MHz

1990MHz1990MHz

Data Voice

A D B E F C A D B E F C

15 51010 1515151515 555 55

Licensed Licensed

Unlicensed

0

Channel Numbers 299

300

400

699700

800

900

1199 0

299300

400

699700

800

900

1199



PCS Band B

PCS Band C

PCS Band D

PCS Band E

PCS Band F

1 425

2 450

3 475

4 500

5 525

6 550

7 575

8 600

9 625

10 650

11 675

1 925

2 950

3 975

4 1000

5 1025

6 1050

7 1075

8 1100

9 1125

10 1150

11 1175

1 325

2 350

3 375

1 725

2 750

3 775

1 825

2 850

3 875



Figure 2-8Overlaying CDMA on the 1900 MHz band

Overlaying CDMA on the 1900 MHz bandIf the 1.25 MHz band of frequencies identified in the drawing as the nominal band is used to support a CDMA channel in some cells, neither the frequencies in this band nor those in the 260 kHz Guard Bands on either side of it can be used in those cells or in the cells in the surrounding Guard Zone (discussed later) for any other purpose.

Frequencies outside this 1.77 MHz band (1.25 MHz nominal plus two 260 kHz guard bands) can be used anywhere, including the central CDMA cells and the surrounding Guard Zone.

Frequencies in this 1.77 MHz band can be used in cells beyond the Guard Zone.


Overlaying CDMA on the 1900 MHz Band

Each CDMA Channel: 1.250 MHz ÷ 50 kHz = 25 channels

Each Guard Band: 260 kHz ÷ 50 kHz = 5.2 = ~5 channels

TOTAL: 1.77 MHz ÷ 50 kHz = 35.4 = ~ 35 channels


Frequency

Po

wer

1.77 MHz

CDMA CARRIER

Just as with the CDMA on AMPS overlay, a GUARD ZONE is also needed

GUARDBAND

GUARDBAND



Figure 2-9Number of voice channels (as AMPS channels are converted to digital)

0

50

100

150

200

1 2 3 4 5 6 7 8 9 10

S1

S4

AMPS

TDMA

13 kbpsCDMA

8 kbpsCDMA

Num

ber

of V

oice

Cha

nnel

s

Number of CDMA Carriers0 1 2 3 4 5 6 7 8 9

200

150

100

50

0



Figure 2-10CDMA frequency channel assignment at 800 MHz (Cellular)

CDMA frequency channel assignment at 800 MHz (Cellular) 2The following table is an example of CDMA channel allocation, in chronological order, which allows maximum CDMA channel packing.

Note: Side A requires frequency coordination with non-cellular interferers. Side B requires frequency coordination with A-side carrier.

493 BTAs (Basic Trading Areas) are grouped into 51 MTAs (Metropolitan Trading Areas).

Order Side A Side B1 283 3842 242 4253 201 4664 160 5075 119 5486 78 5897 37 6308 1019a 777a

9 691 736b

IS-95 RECOMMENDS TO START CDMA DEPLOYMENTWITH EITHER THE PRIMARY OR THE SECONDARY CHANNEL

1

334

667

991

1023

333

666

715

799

716

ChannelNumbers

A Band B Band A’A” B’

1013 31 73 115 157 199 241 283 384 426 468 510 552 594 636 691 777

CDMA A-Band Carriers CDMA B-Band Carriers

8 7 6 5 4 3 2 1 1 2 3 4 5 6 7 9 9 8

* **** Requires frequency coordination with

non-cellular interferers

** Requires frequency coordination with A-band carrier

A Band Primary Channel 283A Band Secondary Channel 691

B Band Primary Channel 384B Band Secondary Channel 777

IS-95 RECOMMENDS TO START CDMA DEPLOYMENTWITH EITHER THE PRIMARY OR THE SECONDARY CHANNEL

1

334

667

991

1023

333

666

715

799

716

ChannelNumbers

A Band B Band A’A” B’

1013 31 73 115 157 199 241 283 384 426 468 510 552 594 636 691 777

CDMA A-Band Carriers CDMA B-Band Carriers

8 7 6 5 4 3 2 1 1 2 3 4 5 6 7 9 9 8

* **** Requires frequency coordination with

non-cellular interferers

** Requires frequency coordination with A-band carrier

A Band Primary Channel 283A Band Secondary Channel 691

B Band Primary Channel 384B Band Secondary Channel 777



Figure 2-11CDMA frequency clearing: A-band (N=7 reuse pattern)

Figure 2-12CDMA frequency clearing: B-band (N=7 reuse pattern)

■ To deploy a CDMA carrier centered on AMPS/TDMA Channel 283, AMPS/TDMA Channels 254 through 312, inclusive, must be cleared from the CDMA coverage area

■ The CDMA channel is implemented, centered on AMPS/TDMA Channel 283. The first usable AMPS/TDMA Channels (outside of the Guard Zone) are Channels 253 and 313

333 332 331 330 329 328 327 326 325 324 323 322 321 320 319 318 317 316 315 314 313

312 311 310 309 308 307 306 305 304 303 302 301 300 299 298 297 296 295 294 293 292

291 290 289 288 287 286 285 284 283 282 281 280 279 278 277 276 275 274 273 272 271

270 269 268 267 266 265 264 263 262 261 260 259 258 257 256 255 254 253 252 251 250249 248 247 246 245 244 243 242 241 240 239 238 237 236 235 234 233 232 231 230 229

228 227 226 225 224 223 222 221 220 219 218 217 216 215 214 213 212 211 210 209 208

207 206 205 204 203 202 201 200 199 198 197 196 195 194 193 192 191 190 189 188 187

186 185 184 183 182 181 180 179 178 177 176 175 174 173 172 171 170 169 168 167 166

165 164 163 162 161 160 159 158 157 156 155 154 153 152 151 150 149 148 147 146 145144 143 142 141 140 139 138 137 136 135 134 133 132 131 130 129 128 127 126 125 124

123 122 121 120 119 118 117 116 115 114 113 112 111 110 109 108 107 106 105 104 103

102 101 100 99 98 97 96 95 94 93 92 91 90 89 88 87 86 85 84 83 82

81 80 79 78 77 76 75 74 73 72 71 70 69 68 67 66 65 64 63 62 61

60 59 58 57 56 55 54 53 52 51 50 49 48 47 46 45 44 43 42 41 40

39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 1918 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1

1A 2A 3A 4A 5A 6A 7A 1B 2B 3B 4B 5B 6B 7B 1C 2C 3C 4C 5C 6C 7C

α β γ

N = 7

α β γ1A 2A 3A 4A 5A 6A 7A 1B 2B 3B 4B 5B 6B 7B 1C 2C 3C 4C 5C 6C 7C

334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354

355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375

376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396

397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417

418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438

439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459

460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480

481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501

502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522

523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564

565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585

586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606

607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627

628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648

649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666

N = 7■ To deploy a CDMA

carrier centered onAMPS/TDMA Channel384, AMPS/TDMAChannels 355 through413, inclusive, must becleared from the CDMAcoverage area

■ The CDMA channel isimplemented, centeredon AMPS/TDMA Channel384. The first usableAMPS/TDMA Channels(outside of the GuardZone) are Channels 354and 414



Figure 2-13Overlay guard zone deployment

Figure 2-14Other technologies: avoiding interference

AMPS Only Cellsapprox 19 channels per sector

CDMA & AMPS Cellsapprox 16 channels per sector

one CDMA channel/carrier/frequency

( 42 + 9 + 9 ) ÷ 21 = 2.8 = ~3 AMPS channels must be cleared persector in the CDMA & AMPS area and in the Guard Zone to make

room for the first CDMA channel/carrier/frequency

The Guard Zones are needed between CDMA and other systemsbecause CDMA increases the noise floor for those systems

AMPS Only Cells (GUARD ZONE)approx 16 channels per cell

■ In conventional radiotechnologies, the desired signalmust be strong enough tooverride any interference

■ AMPS, TDMA and GSM dependon physical distance separationto keep interference at lowlevels

■ Co-channel users are kept at asafe distance by carefulfrequency planning

■ Nearby users and cells mustuse different frequencies toavoid interference

2

3

4

5 6

7

4

6

4

7 2

7

2

5

3

5

36

1

1

1

1

1

1

1

AMPS-TDMA-GSM

Figure of Merit: C/I(carrier/interference ratio)

AMPS: +17 dBTDMA: +14 to 17 dBGSM: +12 to 14 dB



Figure 2-15CDMA: using a new dimension

■ All CDMA users occupy the samefrequency at the same time!Frequency and time are not usedas discriminators

■ CDMA operates by usingCODES to discriminate betweenusers

■ CDMA interference comes mainlyfrom nearby users

■ Each user is a small voice in aroaring crowd -- but with auniquely recoverable code

■ Transmit power on all users mustbe tightly controlled so theirsignals reach the base station atthe same signal level

Figure of Merit: Ec/Io, Eb/No

(energy per chip [bit] /interference [noise] spectral density)

CDMA: Ec/Io -17 to -2 dBCDMA: Eb/No ~+6 dB



What happens during a call? 2Figure 2-16The network view

Figure 2-17The handset view

ConvolutionalEncoder

R=1/2 K=9Symbol

RepetitionPacket

Routing

9600 bps4800 bps2400 bps1200 bps

T-164 kbsPCM

BlockInterleaver

19200 sps9600 sps4800 sps2400 sps

19.2Ksps

Long CodePN

Generator

1.2288Mcps

Decimator÷64

Decimator÷64

MUX

DataScrambling

WalshCode

Wt

User AddressMask (ESN)

19.2Ksps

I PN+∆t

Q PN +∆t

Up-Conversion

Σ Iother users

Σ Qother users

Correlator

Correlator

Correlator

Correlator

Combiner

BlockDe-Interleaver

ViterbiDecoder

VoiceCoding

SwitchingPacket

Routing

Switching

T-164 kbsPCM

BCN

BCNT-1

Unch.

T-1Unch.

PN +∆t

CDSUCDSU

PacketRouting

VoiceCoding CDSU

PacketRouting

CDSU

BCN

BCN BCN

800 Hz

HPA

LNA

IF RF RF

19.2Ksps

1.2288Mcps

BTU/STU

RFIFDe-

modulation

1.2288Mcps

IFModulation

DownConversion

MTX BSC BTS

Power ControlDecision


R=1/2 K=9Symbol

RepetitionPacket

Routing

9600 bps4800 bps2400 bps1200 bps

T-164 kbsPCM

BlockInterleaver

19200 sps9600 sps4800 sps2400 sps

19.2Ksps

Long CodePN

Generator

1.2288Mcps

Decimator÷64

Decimator÷64

MUX

DataScrambling

WalshCode

Wt

User AddressMask (ESN)

19.2Ksps

I PN+∆t

Q PN +∆t

Up-Conversion

Σ Iother users

Σ Qother users

Correlator

Correlator

Correlator

Correlator

Combiner

BlockDe-Interleaver

ViterbiDecoder

VoiceCoding

SwitchingPacket

Routing

Switching

T-164 kbsPCM

BCN

BCNT-1

Unch.

T-1Unch.

PN +∆t

CDSUCDSU

PacketRouting

VoiceCoding CDSU

PacketRouting

CDSU

BCN

BCN BCN

800 Hz

HPA

LNA

IF RF RF

19.2Ksps

1.2288Mcps

BTU/STU

RFIFDe-

modulation

1.2288Mcps

IFModulation

DownConversion

MTX BSC BTS

Power ControlDecision

Duplexer &Bandpass

FiltersIF

BPFMixerLNA

LocalOscillator

(Synthesized)

Traffic Correlator

PN Generator Walsh Generator

Traffic Correlator


Traffic Correlator


Vocoder

Search Correlator (Pilots)

PN Generator Walsh =0 CPU &Control

Algorithms

VocoderConv. Encoder& Symbol Rep.

BlockInterleaver

OrthogonalModulator

Data BurstRandomizer

Direct Seq.Spreading

QuadratureSpreading

BasebandFiltering

IF Modulator

PowerAmplifier

IF

AntennaReceiver

Transmitter

voice bits

voice bits

audio

audio

symbols

chips

Mixer

chips symbols symbols

RF IF

RF

RF

LO

LO

IF

Open Loop Pwr Control

messagebits

message bitsLONG CODE Generator

IF

IF Transmit Gain Adjust: Closed Loop Pwr Control

IF Σ ViterbiDecoder

Duplexer &Bandpass

FiltersIF

BPFMixerMixerLNALNA

LocalOscillator

(Synthesized)

Traffic Correlator


Traffic Correlator


Traffic Correlator


Vocoder

Search Correlator (Pilots)

PN Generator Walsh =0 CPU &Control

Algorithms

VocoderConv. Encoder& Symbol Rep.

BlockInterleaver

OrthogonalModulator

Data BurstRandomizer

Direct Seq.Spreading

QuadratureSpreading

BasebandFiltering

IF Modulator

PowerAmplifier

IF

AntennaReceiver

Transmitter

voice bits

voice bits

audio

audio

symbols

chips

MixerMixer

chips symbols symbols

RF IF

RF

RF

LO

LO

IF

Open Loop Pwr Control

messagebits

message bitsLONG CODE Generator

IF

IF Transmit Gain Adjust: Closed Loop Pwr Control

IF Σ ViterbiDecoder



Exercise 2-1 Lesson review


1. What is the C/I ratio (in decibels) for GSM users?

2. Processing Gain (W/R) is

— the rate of bit rate to bandwidth

— the rate of bandwidth to bit rate

— all of the above

— none of the above

3. Frame Error Rate (FER) is a better performance measurement than Bit Error Rate (BER)?

4. All CDMA RF carriers

— are 1.25 MHz wide

— can serve ~ 22 users with a 8kb vocoder

— can serve ~17 users with a 13 kb vocoder

— all of the above


5. As the number of voice channels increase, the number of AMPS carriers decrease.

6. Long Code PN generation occurs in the

— MTX

— BSC

— BTS *




Lesson 3 CDMA forward channelsObjectives 3


• describe the process of vocoding

• identify the four Forward Traffic Channels and their function

• describe how the four Forward Traffic Channels are generated and the main modulation parameters associated with them

• demonstrate how spreading and de-spreading work in a composite signal made of three different bit streams

3-2 CDMA forward channels Nortel Networks Confidential


Figure 3-1CDMA forward traffic channels

Forward channel coding process 3A Forward Traffic Channel is identified by:

• its CDMA RF carrier frequency

• the unique Short Code PN Offset of the sector

• the unique Walsh Code of the user

■ Used for the transmission of user and signaling information to aspecific mobile station during a call

■ Maximum number of traffic channels: 64 minus one Pilot channel, oneSync channel, and 1 through 7 Paging channels

• This leaves each CDMA frequency with at least 55 traffic channels• Unused paging channels can provide up to 6 additional channels• Realistic loading will typically be about 17 subscribers when using

the 13 kb vocoder (22 when using the 8 kb vocoder)

Forward Traffic Channel

Sync

Paging



Pilot

Σ

CDMA Cell Site

Nortel Networks Confidential CDMA forward channels 3-3


Figure 3-2CDMA code channels in the forward direction

Pilot Walsh 0

Walsh 19

Paging Walsh 1

Walsh 6

Walsh 11

Walsh 20

Sync Walsh 32

Walsh 42

Walsh 37

Walsh 41

Walsh 56

Walsh 60

Walsh 55

■ PILOT: WALSH CODE 0• The Pilot is a “structural beacon” which

does not contain a character stream. It isa timing source used in system acquisitionand as a measurement device duringhandoffs

■ SYNC: WALSH CODE 32• This carries a data stream of system

identification and parameter informationused by mobiles during system acquisition

■ PAGING: WALSH CODES 1 up to 7• There can be from one to seven paging

channels as determined by capacityneeds. They carry pages, systemparameters information, and call setuporders

■ TRAFFIC: any remaining WALSH codes• The traffic channels are assigned to

individual users to carry call traffic. Allremaining Walsh codes are available,subject to overall capacity limited by noise



Figure 3-3Coding process on CDMA forward code channels

Each user is assigned one of the 64 Walsh Codes and their traffic is mixed with the Walsh code to establish a dedicated code channel.

Each user’s Long code is applied incidentally for data scrambling.

All user’s code signals are then analog-summed with the pilot, sync and paging signals to produce one composite waveform.

The composite waveform is the combined with the I and Q Short PN Code sequences using a specific offset to uniquely identify this cell sector.

BTS (1 sector)MTX BSC

FECWalsh #1

Sync FECWalsh #32

Walsh #0

FECWalsh #12

FECWalsh #23

FECWalsh #27

FECWalsh #44

Pilot

Paging

Vocoder

Vocoder

Vocoder

Vocoder

more moremore

Trans-mitter,

Sector X

Σ

I Q

Short PN CodePN Offset 246

A Forward Channelis identified by:

❖ its CDMA RFcarrier Frequency

❖ the unique ShortCode PN Offset ofthe sector

❖ the unique WalshCode of the user

CDMAFrequency



Figure 3-4Digital Stream 0 (DS0)

Pulse Code Modulation (PCM) is the most common method of encoding an analog voice signal into a digital bit stream.

SamplingTo digitize an analog signal, it must first be sampled at very precise intervals of time to measure its amplitude. The sampling period is dictated by the application of Nyquist’s theorem, which states that the sampling rate must be at least double the highest frequency to be transmitted if the original signal is to be recovered from these samples with minimum error.

The highest frequency component in an analog speech signal is less than 4 KHz; therefore, the sampling rate is rounded up to eight thousand times per second.

QuantizingOnly discrete values are applicable in a digital system; therefore, each sample is allocated an integer value that represents the signal level. Eight bits are commonly allocated to represent the samples. Eight bits times 8000 samples per second results in 64,000 bits per second, the familiar rate of a DS0 bit stream.

tt

0

1

2

3

4

56

87

91011

12

13

14

15

16

4

16

1

3

15

8

34

8

t

0

1

2

3

4

56

87

91011

12

13

14

15

16

8

15

3

1

3

16

4

4

8

8

15

3

1

3

16

4

4

8

t

8

15

3

1

3

16

4

4

8

8

15

3

1

3

16

4

4

8

64 kbs

Analog Voice Signal Sampling Quantizing

Signal Regeneration



Linear quantization produces proportionally more error on smaller amplitude signals; on the other hand, large amplitude signals are better able to mask the effects of noise. This imbalance can be improved by using a non-linear form of quantization which acts as level compression during encoding and level expansion during decoding. This technique is referred to as “companding” (from compressing and expanding).

Simplified vocoder functions 3The Codebook stores a collection of arbitrary waveform segments (a sort of digitized vocal clip art collection) in digital form. Within the 20ms sample time, the vocoder–through approximation based upon previous samples–approximates as closely as possible a code representation of the sample signal.

The Pitch Filter can be thought of as modeling the periodic pulse train coming from the vocal cords during voiced speech.

The Formant Filter models the characteristics of the vocal tract. It has resonant frequencies near the resonant frequencies of the original speech caused by the vocal tract filtering.

Figure 3-5Variable rate vocoding & multiplexing (Traffic Channels only)

■ Vocoders compress speech,reduce bit rate

■ CDMA uses a superiorVariable Rate Vocoder• full rate during speech• low rates in speech pauses• increased capacity• more natural sound

■ Voice, signaling, and usersecondary data may bemixed in CDMA frames

DSP QCELP VOCODER

Codebook

PitchFilter

FormantFilter

Coded Result Feed-back

20ms Sample

Rate Set 2 Frame Sizesbits

Full Rate Frame

1/2 Rate Frame

1/4 Rt.

1/836

72

144

288

Frame Contents: can be a mixture ofVoice Signaling Secondary



Digital Signal Processors (DSPs)—Special purpose microprocessors designed specifically for high-speed signal processing applications such as speech coding, signaling tone-generation and detection, and speech synthesis.

VSELP—Vector Sum Excited Linear Predictive encoding.

QCELP—Qualcomm Code Excited Linear Predictive encoding.

Figure 3-6Converting bits into symbols

BITS—The 0s and 1s representing the original data in one frame of a particular forward or reverse code channel are called bits.

SYMBOLS—The 0s and 1s resulting of the addition of Forward Error Correction redundancy to the original bits are referred to as symbols. There are two types of symbols:

• CODE SYMBOLS, which are the 0s and 1s resulting of processing the data bits in a frame through the convolutional encoder.

• MODULATION SYMBOLS, which are the smallest independent units of information spread and modulated with a Walsh code.

■ The bits in a 20 ms traffic frame may include oneor more of the following

• voice information (from the vocoder)• signaling information• secondary traffic information

■ When Forward Error Correction algorithms areapplied to these information bits, the resulting 0sand 1s are called symbols

• bits and symbols are related in a complexmany-to-many fashion

− the information in one bit is distributed amongmany symbols, and one symbol carries someof the information of many bits

• all forward traffic frames contain 384 symbols• all reverse traffic frames contain 576 symbols

Bits

Symbols

Forward ErrorCorrection



— In the forward code channels each repetition of a code symbol is a modulation symbol. (There are 16 * 24 = 384 code symbols in one Forward Traffic Channel frame.)

— In the reverse code channels each group of six successive code symbols in a “power control group” is a code symbol. (There are 16 * 36 = 576 code symbols in one Reverse Traffic Channel frame. Each one of these sixteen 36-symbol groups is called a power control group.)

Figure 3-7Spreading symbols into chips

A Walsh Code is a sequence of 64 0s and/or 1s with a particular pattern.

A Walsh Function is a continuous supply of a Walsh Code pattern generated 19200 times per second. That is, 1.2288 million of 0s and/or 1s generated every second.

CHIPS—The 0s and 1s resulting from spreading the modulation symbols with Walsh Codes are referred to as chips.

The 0s and 1s in a Walsh Function, as well as those in the short and long PN sequences (all of which are generated at a rate of 1.2288 Mcps), are also called “chips”.

■ Symbols are converted into special 64-chippatterns for transmission

• there are 64 such patterns called “Walshcodes”

• in the forward link, just one of these patterns isassigned to each user’s stream of symbols(code channel)

− each ‘0’ symbol is replaced by the selectedpattern (Walsh code)

− each ‘1’ symbol is replaced by the logicalnegation of the selected pattern

• in the reverse link, all the 64 patterns (but nottheir logical negations) are used in every codechannel

− each group of six symbols is interpreted asa binary value in the 0-63 range andreplaced by the corresponding Walsh code

Symbols

Chips

Coding andSpreading



Spreading the Modulation Symbols raises the rate to 1.2288 Mcps in the Paging and Forward Traffic Channels, and to 307.2 kcps in the Access and Reverse Traffic Channels.

The chip rate in the Access and Reverse Traffic Channels is later raised to the standard 1.2288 Mcps by direct sequence spreading with the long PN code offset by the appropriate mask.

Figure 3-8Reversing the process

Signal regenerationIn practice, any signal that is reconstructed from digitized samples will suffer from the approximations made during this quantization process. Such errors produce quantization noise.

■ To revert the process, first the symbols arerecovered as follows• in the forward direction, the mobile station correlates

the received signal with the selected Walsh codepattern (integrating the power over 64 chips); a perfectmatch corresponds to a ‘0’ symbol; a perfect no-matchcorresponds to a ‘1’ symbol; for anything betweenthese extremes, the mobile station guesses based onwhich case the partial match resembles closer

• in the reverse direction, the BTS matches thereceived signal with each possible Walsh code andselects the pattern that produces the highest degree ofcorrelation as the representation of the last group ofsix symbols sent

■ When all the symbols for a 20 millisecond framehave been recovered, the Viterbi decoder is used toguess the block of bits (frame) that most probablycorresponds to this block of symbols

■ Then, the CRC of this frame is calculated todetermine if the guess was successful; if not, theframe is discarded (or “erased”)

Symbols

Chips

Despreading(integraton)

Bits

ViterbiDecoder



Figure 3-9Forward traffic channels: Vocoding


Forward Traffic Channels: Vocoding

■ Vocoding reduces the bit rate needed to represent speech■ Output is 20 ms frames at fixed rates:

Full Rate, 1/2 Rate , 1/4 Rate , 1/8 Rate, & Blank■ CRC is added to all the frames for the 13 kb vocoder, but

only to the Full and 1/2 rate frames for the 8 kb vocoder■ CRC is not added to the lower rate frames in the 8 kb

vocoder but that is ok because they consist mostly of background noise and have a higher processing gain

■ Current vocoder rates are 13 kb, 8 kb, and 8 k EVRC (Enhanced Variable Rate Coder)

To theConvolutional

Encoder

20 ms slices(1280 bits)

Variable RateVoice Coding

Add CRCAdd 8 bit

Encoder Tail

64 kbpsFrom MTX

ConvolutionalEncoding

Code Sym bolRepetition

BlockInterleaving

Data Scram bling

Power ControlSubchannelOrthogonalSpreading

QuadratureSpreadingBasebandFiltering

VocoderProcessing

Baseband Trafficto RF Section

PCM Voice

BSC

BTS

(SymbolPuncturing)



Figure 3-10Variable rate vocoder

Speech coding takes advantage of the fact that most typical voice conversations consist of better than 50% dead (or idle) time. Thus, it makes sense to compress voice traffic and send only intelligence, thereby increasing capacity. As shown later, CDMA also takes advantage of this to decrease the overall required user power.

The average duty cycle for each speaker in a conversation is estimated at about 35% to 40% of the time.

A-to-DCONVERTER

64 kbps

VOCODER

“Codebook” Instruction(< 64 kbps)

■ Speech coding algorithms (digital compression) are necessary toincrease cellular system capacity

■ Coding must also ensure reasonable fidelity, i.e., a minimum levelof quality as perceived by the user

■ Coding can be performed in a variety of ways (ex. waveform, timeor frequency domain)

■ Vocoders transmit parameters which control “reproduction” ofvoice instead of the explicit, point-by-point waveform description



Figure 3-11Wireless data service

The Service Options that are required to support Wireless Data Services includes Asynchronous Data and Group 3 Facsimile for both Rate Set 1 and Rate Set 2. The Service Options are responsible for receiving forward data traffic from the DSP, which is received from the InterWorking Function (IWF), and passing it to the Traffic Processing software, on the Selector Card, to be multiplexed and sent down to the BTS(s). The Service Options are also responsible for receiving reverse data traffic frames from the Traffic Processing software and passing them to the DSP for IS-99 Radio Link Protocol (RLP) processing and sending on to the IWF.

As part of the Wireless Data Services the Service Options, on the SBS, necessary to support data calls for both Rate Set 1 and 2. Data calls include Fax or Asynchronous Data Service Options. Specific service options that are supported by this feature are:

• Asynchronous Data

9600bps at Rate Set 1 or 14400bps at Rate Set 2

• Group 3 Facsimile

9600bps at Rate Set 1 or 14400bps at Rate Set 2

The Data Service Options are responsible for passing traffic between the Traffic Processing software and the DSP, which all reside on Selector Card.

■ The SBS selector Card Service Options include :• Asynchronous Data (9600 bps) at Rate Set 1 and (14400 bps) at Rate Set 2• Group 3 Facsimile (9600 bps) at Rate Set 1 and (14400 bps) at Rate Set 2

Traffic Processing

Data ServiceOption

IWF

BTS

SBS SelectorSBS Selector

IS-95 RLP

Traffic Frames

IWF

TrafficDSP



The Traffic Processing software is responsible for the multiplexing/demultiplexing of traffic and signalling information to and from the BTS(s)/Mobile. The Traffic Processing software sends and receives the data traffic frames to/from the Data Service Option.

The Data Service Option is responsible for providing the Traffic Processing software forward data traffic frames every 20 milliseconds and the DSP reverse data traffic frame every 20 milliseconds. The Data Service Option also provides an interface for communications to/from the DSP.

The DSP is loaded with software, during call setup, that provides the RLP portion of the IS-99 protocol stack that relays the data traffic to the IWF.

Figure 3-12Forward traffic channel generation

Vocoding in the forward direction takes place at the BSC, in one of the Cell Site Modem (CSM) chips. The reverse process will take place at the mobile station in the Mobile Station Modem (MSM) chip. In the rest of this and the following lesson, we will be concerned with what happens in the Channel Element of the BTS that has been assigned to handle one particular conversation.

In this lesson, we deal with the convolutional coding, repetition, and block interleaving of the 20 ms slices of conversation delivered by the vocoder which result in the creation of the forward traffic frames.

GainControl

BasebandFilter

BasebandFilter

I PN

Q PN

1.2288Mcps

WalshFunction

BlockInterleaving

R = 1/2, K = 9Convolutional

Encoding &Repetition

19.2Ksps

19.2 Ksps

DecimatorLong PN CodeGenerator

Scrambling

User AddressMask

(ESN-Based)

1.2288Mcps

19.2Ksps

9600 bps4800 bps2400 bps1200 bps

or14400 bps7200 bps3600 bps1800 bps

(traffic frames)

PowerControl

Bit

Decimator

800 Hz

MUX

SymbolPuncturing(13 Kb only)

28.8Ksps

bits symbols chips

CHANNEL ELEMENT



In lesson 4, we discuss all the other transformations experimented by the frames until the signal is ready to go to the antenna.

As we discuss all these transformations suffered by each 20 ms slice of conversation in the BTS, we will also explain how these actions are reversed at the mobile station.

Figure 3-13Forward traffic channel frame structure

Error checking is not performed on the 2400 bps and 1200 bps frames because they consist largely of background noise. Additionally, the lower rates benefit from a higher processing gain.

Notice that the lowest two rates in Rate Set 1 have no CRC because they benefit from a higher processing gain. Why is it so?

Remember the definition of processing gain: W/R. The same bandwidth 1.2288 Mbps is used to transmit a lower rate 4800 bps, 2400 bps, and 1200 bps; therefore, the processing gain (W/R) is twice, four times, and eight times higher, respectively.

TransmissionRate (bps) Total Reserved Information CRC Tail Bits

9600 192 — 172 12 8

4800 96 — 80 8 8

2400 48 — 40 — 8

1200 24 — 16 — 8

14400 288 1 267 12 8

7200 144 1 125 10 8

3600 72 1 55 8 8

1800 36 1 21 6 8

1

2

Number of Bits per Frame (20 ms)RateSet



Figure 3-14Convolutional encoding and symbol repetition


Convolutional Encoding andSymbol Repetition

■ Convolutional encoding• Is a means of error detection/correction • Results in 2 code symbols (or more, depending on the

“R” constant) output for each bit input■ Symbol repetition maintains a constant 19.2 Ksps output to

be fed into the block interleaver• Also allows for reduction in transmit power• Reduces overall noise and increases capacity


Code SymbolRepetition

BlockInterleaving

Data Scrambling

Power ControlSubchannelOrthogonalSpreadingQuadratureSpreadingBasebandFiltering

VocoderProcessing


PCM Voice

(SymbolPuncturing)

Variable RateOutput fromthe Vocoder


R=1/2 K=9

SymbolRepetition

19.2 kspsto Block

Interleaver

14.4 kbps7.2 kbps3.6 kbps1.8 kbps

28.8 ksps14.4 ksps

7.2 ksps3.6 ksps

9.6 kbps4.8 kbps2.4 kbps1.2 Kbps

19.2 ksps9.6 ksps4.8 ksps2.4 ksps

28.8 kspsto Symbol Puncturing

8 kb

13 kb

bits codesymbols

modulationsymbols



Figure 3-15A very simple convolutional encoder

Only a specified number of bits are processed by the convolutional encoder at any given time. As a bit enters the shift register, the bit stream is staggered by one. The length of the shift register plus one is referred to as the constraint length of the convolutional encoder.

In this example, two encoding processes are active. The top process encodes three bits while the bottom process encodes two. Therefor, for each encoded bit stream two code symbols are generated: one by the top process and another by the bottom process.

Once the bit has been shifted through the register, it is discarded. If one of these code symbols is corrupted, only 1/2 of 1/constraint length of the information in k contiguous data bits is lost, giving us a better chance to correct the error at the receiving end than if a whole data bit had being altered. On the other hand, if “2k” contiguous code symbols in the stream were corrupted, all the information needed to reconstruct one data bit would be compromised (unless some preventive measured, called “block interleaving” is introduced).


A Very Simple Convolutional Encoder

+

+

1011000



The problem is: how can we figure out the original sequence of data bits if all we have is the code symbol output?

Figure 3-16Rate 1/2, K=9 convolutional encoding

A convolutional encoder accepts data bits in and outputs code symbols. With each clock cycle, a new data bit is shifted into block 1 of the register, and the data bit previously in the last block is dumped. The inputs to the various taps are added (modulo 2) to produce two or more symbols out each clock cycle; these symbols are represented by C0, C1, etc.

Since the symbols generated with each clock cycle are derived from the values of a new bit being input and all the current data bits occupying the shift register during a given interval, a certain level of predictability, hence corrigibility, can be obtained.

To avoid confusion, the 0’s and 1’s resulting of the combination of multiple original data bits in a convolutional encoder are no longer called “data bits” but “code symbols.”

■ Symbols generated as the information bits transit through the encoder, arerelated to all the bits currently in the register

■ Each information bit contributes to multiple generated symbols■ This pattern of inter-relationships helps detect and correct errors■ The length of shift register plus 1 is called the “constraint length” of the

convolutional encoder (K=9 in this case)• The longer the register, the better this scheme can correct bursty errors• Reduces power required to achieve same accuracy as without coding

■ Here, two symbols are generated for every bit input (Rate 1/2)

Code SymbolOutput

1 2 3 4 5 6 7 8

g0

g1

c0

c1

DataBit

Input

(Data Bit isdiscarded)

Code SymbolOutput



Figure 3-17Symbol repetition and power reduction

The ability to reduce the energy level of lower rate symbols is an essential element to CDMA system design. Recalling that the primary source of interference for any user is other users on the same channel, the ability to minimize interferer noise through power reduction increases overall system capacity.

As the above chart shows, information bits at each data rate retain the same level of power. It is simply spread across more symbols at the lower rates. When processed at the receiver, the repeated symbols, when integrated, add up to the same power level as a full rate single symbol.


Symbol Repetition and Power Reduction

■ Symbol repetition provides a constant rate to the block interleaver■ Lower rates symbols are sent at reduced power levels■ The energy per bit across all rates is identical when integrated■ Overall signal power requirement (thus noise) is reduced

Data Rate(bps)

Energy perModulation Symbol

14000 / 9600

72000 / 4800

3600 / 2400

1800 / 1200

E =E /2s b

E =E /4s b

E =E /8s b

E =E /16s b

Full Energy

1/2 Energy

1/4 Energy

1/8 Energy

MATHHAMMER



Figure 3-18Symbol puncturing – Rate Set 2 (13 kb vocoder)

For the Forward Traffic Channel Rate Set 2, an effective code rate of 3/4 is achieved by puncturing two out of every symbols after symbol repetition. The effective code rate is the rate of the convolutional encoder (1/2) divided by the puncturing rate (4/6).

The puncturing pattern is 110101, where a “0” means that the symbol is deleted and the most significant bit in the pattern corresponds to the first symbol in the six symbol group.


Symbol PuncturingRate Set 2 (13 kbps Vocoder)

■ Symbol repetition maintains a constant 28.8 ksps output to puncturing section

■ Symbol puncturing deletes 2 of every 6 inputs based on a six-bit pattern

■ Unrepeated symbols for 28.8 ksps frames are also deletedConvolutional decoder in mobile station will correct these

purposeful errors ■ Puncturing provides a constant 19.2 Ksps input to interleaver

just like in rate set 1This allows all other functions to remain exactly the same

PCM Voice



BlockInterleaving

Data Scrambling


VocoderProcessing


(SymbolPuncturing)

FromR=1/2 K=9

Convolutional Encoder

SymbolPuncturing

to the BlockInterleaver

SymbolRepetition

28.8 ksps28.8 ksps14.4 ksps

7.2 ksps3.6 ksps

19.2 Ksps



Figure 3-19Block interleaving

Fast fading can destroy segments of data. By resequencing the symbols such that adjacent symbols have no direct relationship to each other, the effects of fast fading can be diminished. This is so because the effect of a fade, when the symbols are reordered at the receiver, is spread across more bits with reduced severity.

■ 20 ms symbol blocks are sequentially reordered■ Combats the effects of fast fading■ Separates repeated symbols at 4800 bps and

below• Improves survivability of symbol data• “Spreads” the effect of bursty interference

19.2 kspsFrom Coding& SymbolRepetition

Input Array(Normal

Sequence)24 X 16

Output Array(ReorderedSequence)

24 X 16To DataScramblingFunction

PCM Voice



BlockInterleaving

Data Scrambling


VocoderProcessing

Baseband Traffic to RF Section

(SymbolPuncturing)



Figure 3-209600 bps block interleaver (input array)

The Paging Channels and the Forward Traffic Channels use identical block interleavers, spanning 20 ms, which is equivalent to 384 modulation symbols at the rate of 19200 symbols per second.

The input (array write) and output (array read) symbol sequence are shown in the block interleaver figures. These tables are read down by columns from left to right. In these tables, symbols with the same number denote repeated symbols.

■ The 384 modulation symbols in a frame are input into a 24 by 16block interleaver array (read down by columns, from left to right)

■ The array represents a 20 ms interval worth of information

1 25 49 73 97 121 145 169 193 217 241 265 289 313 337 3612 26 50 74 98 122 146 170 194 218 242 266 290 314 338 3623 27 51 75 99 123 147 171 195 219 243 267 291 315 339 3634 28 52 76 100 124 148 172 196 220 244 268 292 316 340 3645 29 53 77 101 125 149 173 197 221 245 269 293 317 341 3656 30 54 78 102 126 150 174 198 222 246 270 294 318 342 3667 31 55 79 103 127 151 175 199 223 247 271 295 319 343 3678 32 56 80 104 128 152 176 200 224 248 272 296 320 344 3689 33 57 81 105 129 153 177 201 225 249 273 297 321 345 369

10 34 58 82 106 130 154 178 202 226 250 274 298 322 346 37011 35 59 83 107 131 155 179 203 227 251 275 299 323 347 37112 36 60 84 108 132 156 180 204 228 252 276 300 324 348 37213 37 61 85 109 133 157 181 205 229 253 277 301 325 349 37314 38 62 86 110 134 158 182 206 230 254 278 302 326 350 37415 39 63 87 111 135 159 183 207 231 255 279 303 327 351 37516 40 64 88 112 136 160 184 208 232 256 280 304 328 352 37617 41 65 89 113 137 161 185 209 233 257 281 305 329 353 37718 42 66 90 114 138 162 186 210 234 258 282 306 330 354 37819 43 67 91 115 139 163 187 211 235 259 283 307 331 355 37920 44 68 92 116 140 164 188 212 236 260 284 308 332 356 38021 45 69 93 117 141 165 189 213 237 261 285 309 333 357 38122 46 70 94 118 142 166 190 214 238 262 286 310 334 358 38223 47 71 95 119 143 167 191 215 239 263 287 311 335 359 38324 48 72 96 120 144 168 192 216 240 264 288 312 336 360 384

�� %QNWOPU

��TQYU



Figure 3-219600 bps block interleaver (output array)

■ This 24 by 16 array (read down by columns, from left to right) indicates theorder in which the symbols are output from the block interleaver

■ The effect of bursty errors during transmission is minimized (the 2kcontiguous symbols containing the information to restore one data bithave been separated)

1 9 5 13 3 11 7 15 2 10 6 14 4 12 8 1665 73 69 77 67 75 71 79 66 74 70 78 68 76 72 80

129 137 133 141 131 139 135 143 130 138 134 142 132 140 136 144193 201 197 205 195 203 199 207 194 202 198 206 196 204 200 208257 265 261 269 259 267 263 271 258 266 262 270 260 268 264 272321 329 325 333 323 331 327 335 322 330 326 334 324 332 328 33633 41 37 45 35 43 39 47 34 42 38 46 36 44 40 4897 105 101 109 99 107 103 111 98 106 102 110 100 108 104 112

161 169 165 173 163 171 167 175 162 170 166 174 164 172 168 176225 233 229 237 227 235 231 239 226 234 230 238 228 236 232 240289 297 293 301 291 299 295 303 290 298 294 302 292 300 296 304353 361 357 365 355 363 359 367 354 362 358 366 356 364 360 36817 25 21 29 19 27 23 31 18 26 22 30 20 28 24 3281 89 85 93 83 91 87 95 82 90 86 94 84 92 88 96

145 153 149 157 147 155 151 159 146 154 150 158 148 156 152 160209 217 213 221 211 219 215 223 210 218 214 222 212 220 216 224273 281 277 285 275 283 279 287 274 282 278 286 276 284 280 288337 345 341 349 339 347 343 351 338 346 342 350 340 348 344 35249 57 53 61 51 59 55 63 50 58 54 62 52 60 56 64

113 121 117 125 115 123 119 127 114 122 118 126 116 124 120 128177 185 181 189 179 187 183 191 178 186 182 190 180 188 184 192241 249 245 253 243 251 247 255 242 250 246 254 244 252 248 256305 313 309 317 307 315 311 319 306 314 310 318 308 316 312 320369 377 373 381 371 379 375 383 370 378 374 382 372 380 376 384

Assume that aburst of noisedamages allthese bits



Figure 3-229600 bps de-interleaving

1 25 49 73 97 121 145 169 193 217 241 265 289 313 337 3612 26 50 74 98 122 146 170 194 218 242 266 290 314 338 3623 27 51 75 99 123 147 171 195 219 243 267 291 315 339 3634 28 52 76 100 124 148 172 196 220 244 268 292 316 340 3645 29 53 77 101 125 149 173 197 221 245 269 293 317 341 3656 30 54 78 102 126 150 174 198 222 246 270 294 318 342 3667 31 55 79 103 127 151 175 199 223 247 271 295 319 343 3678 32 56 80 104 128 152 176 200 224 248 272 296 320 344 3689 33 57 81 105 129 153 177 201 225 249 273 297 321 345 369

10 34 58 82 106 130 154 178 202 226 250 274 298 322 346 37011 35 59 83 107 131 155 179 203 227 251 275 299 323 347 37112 36 60 84 108 132 156 180 204 228 252 276 300 324 348 37213 37 61 85 109 133 157 181 205 229 253 277 301 325 349 37314 38 62 86 110 134 158 182 206 230 254 278 302 326 350 37415 39 63 87 111 135 159 183 207 231 255 279 303 327 351 37516 40 64 88 112 136 160 184 208 232 256 280 304 328 352 37617 41 65 89 113 137 161 185 209 233 257 281 305 329 353 37718 42 66 90 114 138 162 186 210 234 258 282 306 330 354 37819 43 67 91 115 139 163 187 211 235 259 283 307 331 355 37920 44 68 92 116 140 164 188 212 236 260 284 308 332 356 38021 45 69 93 117 141 165 189 213 237 261 285 309 333 357 38122 46 70 94 118 142 166 190 214 238 262 286 310 334 358 38223 47 71 95 119 143 167 191 215 239 263 287 311 335 359 38324 48 72 96 120 144 168 192 216 240 264 288 312 336 360 384

�� %QNWOPU

��TQYU

Notice how the effect of the burst of noise is spread over the transmitted block



Figure 3-23Forward channel demodulation

The Search Correlator searches for other pilot PN sequence offsets (that is, searching for other cell sites with sufficient signal strength for demodulation). Additionally, it identifies usable multipath signal components and assigns these to available correlators. This function permits the mobile station to combine multipath signals, to improve receive signal quality, and to operate in soft handoff. These signals must be on the same CDMA RF channel.

■ IS-95A/J-STD-008 requires a minimum of four processingelements that can be independently directed

• Three elements must be capable of demodulating multipathcomponents

• One must be a “searcher” that scans and estimates signalstrength at each pilot PN sequence offset

Correlator 1

Correlator 2

Correlator 3

Search Correlator

De-Interleaver Decoder Vocoder SpeechOutput

Com

bine

r

Mobile Receiver



Figure 3-24Putting it all together: CDMA code channels

■ The three spreading codes are used in different ways to create theforward and reverse links

■ A forward channel exists by having a specific Walsh Codeassigned to the user, and a specific PN offset for the sector

■ A reverse channel exists because the mobile uses a specific offsetof the Long PN sequence

BTS

WALSH CODE: Individual UserSHORT PN OFFSET: Sector

LONG CODE OFFSET: individual handset

FORWARD CHANNELS

REVERSE CHANNELS

LONG CODE:Data

Scrambling

WALSH CODES:used as symbols

for robustness

SHORT PN:used at 0 offset

for tracking



Pilot channel 3The pilot channel is a reference channel which the mobile station uses for acquisition, timing, and as phase reference for coherent demodulation. It is transmitted at all times by each base station on each active CDMA frequency. This signal is tracked continuously by each mobile station.

Unlike the long code sequence, which has a very long interval between repetitions, the pilot sequence is repeated once every 262/3 ms or 75 times every two seconds. Not only does this aid in initial acquisition when the mobile station (for instance) powers up, but also ensures rapid detection of handoff candidates.

Figure 3-25Pilot channel

■ Used by the mobile station for initial system acquisition■ Transmitted constantly by the base station■ The same PN sequences are shared by all base stations

• Each base station is differentiated by a phase offset■ Provides tracking of

• Timing reference• Phase reference

■ Separation by phase provides for extremely high reusewithin one CDMA channel frequency

■ Acquisition by mobile stations is enhanced by• Short duration of Pilot PN sequence• Uncoded nature of pilot signal

■ Facilitates mobile station-directed handoffs• Used to identify handoff candidates• Key factor in performing soft handoffs



Figure 3-26Pilot channel generation

The pilot channel is sent unmodulated, and it is orthogonally spread with Walsh function zero (which ensures it is easily recognized). Quadrature spreading and channel filtering occur exactly as discussed for the forward traffic channel.

Distinct Pilot Channels are identified by a PN offset index (0 through 511 inclusive). This PN offset index specifies the offset value from the zero offset pilot sequence. The PN offset in chips for a given pilot PN sequence equals the offset index multiplied by 64.

The zero offset pilot PN sequence is aligned with the beginning of System Time (January 6, 1980 at 00:00:00 hrs, Universal Coordinated Time) and with every even-second boundary thereafter, referred to the base station transmission time (within ±10µs of the CDMA System Time). The sequence itself fits exactly 75 times within a two-second interval.

The starting point of the short PN code (zero pilot PN offset) is the ‘1’ bit in either the I or the Q sequence following 15 zeroes. The starting point of the long PN code is the ‘1’ bit following 42 zeroes in the sequence. They starting point of the long PN code is aligned with the beginning of system time.

■ The Walsh function zero spreading sequence is applied to the Pilot■ The use of short PN sequence offsets allows for up to 512 distinct

Pilots per CDMA channel■ The PN offset index value (0-511 inclusive) for a given pilot PN

sequence is multiplied by 64 to determine the actual offset• Example: 15 (offset index) x 64 = 960 PN chips• Result: The start of the pilot PN sequence will be delayed

960 chips x 813.8 nanoseconds per chip = 781.25 µs■ The quadrature spreading and baseband filtering (not shown),

which are performed as with all the other forward and reverse codechannels, will be discussed later

GainControl

BasebandFilter

BasebandFilter

I PN

Q PN

1.2288Mcps

WalshFunction 0

PilotChannel(All 0’s)



Figure 3-27Walsh Codes generation

The Walsh codes used in CDMA are based on a Walsh or Hagamard matrix, which is a square matrix with binary elements. The order (number of rows or columns) of a Walsh or Hagamard matrix is always a power of 2.

These matrices are generated by seeding Walsh (1) = W1 = 0 and expanding from there, as shown in the illustration.

Notice that if the Walsh matrixes were generated by seeding Walsh (1) with 1 instead of 0, we would have obtained the sequence of the corresponding logically negated matrixes. The logical negation of a Walsh matrix is another matrix of the same order but with every element logically negated (that is, a new matrix where the roles of ones and zeroes are reversed). A dash above a “W” in the illustration denotes a logically negated matrix.

CDMA uses the Walsh matrix of order 64 (that is, of size 64 x 64). The rows of this matrix are called Walsh codes, and they are indexed 0 through 63 from the top down.

W1 = 0 0 00 1W2 =

0 0 0 00 1 0 10 0 1 10 1 1 0

W4 =

W2 n = Wn Wn

Wn Wn

W1 = 1 1 11 0W2 =

1 1 1 11 0 1 01 1 0 01 0 0 1

W4 =



Figure 3-28CDMA “Short” and “Long” PN codes

The long code is used for data scrambling the Reverse Traffic Channels. Although all mobile and base stations use the same long code generator perfectly synchronized, each mobile station combines the generated long code with a unique mask (based on its unique Electronic Serial Number or ESN) which produces a unique offset. To prevent the possibility of the close correlation of two subscribers due to their ESNs being consecutive, the ESN portion of the mask is permuted in a predefined way.

CDMA uses three PN code sequences: two “short” and one “long”■ The two short PN codes (called “I” and “Q”) are used for quadrature

spreading to differentiate between CDMA partitions (sectors/cells) inthe forward direction

■ The two short codes are generated by 15-bit PN code generators.The generated strings are 215 -1 bits long plus one zero insertedfollowing the longest string of generated zeroes (32,768); and theircycle period is 26.666... milliseconds (or 75 times every 2 seconds).

■ The long PN code is used for spreading and datascrambling/randomization, and to differentiate among mobile stationsin the reverse direction.

■ The long code is generated by a 42-bit PN code generator. Thegenerated string is 242 -1 with no zero inserted (about 4.4 trillion) bitslong; and its cycle period is approximately 41 days, 10 hours, 12minutes and 19.4 seconds.

■ The three CDMA PN codes are synchronized to the beginning ofsystem time (January 6, 1980 at 00:00:00 hours)



Figure 3-29Pilot channel acquisition

■ The mobile station starts generating the I and Q PN short sequencesby itself and correlating them with the received composite signal atevery possible offset

• In less that 15 seconds (typically 2 to 4 seconds) all possibilities (32,768)are checked

• The mobile station remembers the offsets for which it gets the bestcorrelation (where the Ec/I0 is the best)

■ The mobile station locks on the best pilot (at the offset that results inthe best Ec/I0), and identifies the pattern defining the start of the shortsequences (a ‘1’ that follows fifteen consecutive ‘0’s)

■ Now the mobile station is ready to start de-correlating with Walshcode 32 to extract the Sync Channel (next section)

00...01 00...01 00...01 00...01 00...01 00...0100...01

PILOT CHANNEL(Walsh Code 0)



Sync channel 3Once a strong Pilot Channel is located, the mobile station listens to the corresponding Sync Channel for system information. This information, transmitted at a rate of 1200 bps, is contained in the Sync Channel Message which is broken into 26.666... ms frames.

Notice that the duration of the Sync Channel frames coincides with the period of the short PN codes transmitted on the Pilot Channel. Therefore, once the mobile station acquires synchronization with the pilot channel, the synchronization with the Sync Channel is immediately known. This facilitates the acquisition of the Sync Channel by the mobile station.

Figure 3-30Sync channel

■ Used to provide essential systemparameters

■ Used during system acquisition stage■ The bit rate is 1200 bps■ The Sync channel has a frame

duration of 26 2/3 ms• this frame duration matches the

period of repetition of the PN ShortSequences

• this simplifies the acquisition of theSync Channel once the Pilot Channelhas been acquired

■ The Mobile Station re-synchronizesat the end of every call

The Pilot channel carries no data, therefore it has no frames.The Sync channel uses 26 2/3 ms frames.All other forward and reverse code channels use 20 ms frames.

(Acquired Pilot)

Sync Channel



Figure 3-31Frames and messages

■ Logical unit of transmission■ Fixed length

• no need for length info

■ Each frame includes one or moreoverhead bits in addition to the“payload” of information bits

• these overhead bits define thestructure of the frame

■ Logical unit of information■ Variable length

• must include length info

■ A message is broken into smallpieces that can fit in the payloadportion of successive frames

• one frame overhead bit could beused to identify the initialsegment of a message

FRAME MESSAGE

1 0 0 0

MESSAGE

FRAME

FRAME+ + + +

+ +Sync

Traffic



Figure 3-32Sync channel generation

The Sync Channel uses the same convolutional encoding scheme (rate 1/2, k=9) as the Forward Traffic Channel (at 1200 bps) with one difference: The convolutional coder is not reset to all 0’s at the end of each frame (that is, no trail bits are added at the end of each frame).

Each symbol coming out of the convolutional encoder is repeated twice. This, combined with the 1/2 rate of the convolutional encoder, quadruples the 1200 bps bit rate of the Sync Channel into 4800 modulation symbols before entering the block interleaver.

The modulation symbols are interleaved as shown in the following two slides. Notice that as the convolutional coder is not “flushed” at the end of each frame, the last eight bits of a Sync Channel frame influence the first 36 modulation symbols input in the successive interleaver block.

The Sync Channel data is not scrambled with the long PN code, and no power control subchannel is inserted either. It is orthogonally spread with Walsh function 32, which is a series of 32 zeros followed by 32 ones. Quadrature Spreading and Channel Filtering is just as before.

■ There are 32 bits (1200 bps x 0.02666... second) in one Sync Channel frame■ The Rate 1/2 convolutional encoder doubles the bit rate, and the resulting 0s and

1s are now called “code symbols”• there are 64 code symbols in a Sync Channel frame

■ The repetition process doubles the rate again, and each repetition of a codesymbol is now called a “modulation symbol”

• there are 128 modulation symbols in a Sync Channel frame■ Four copies of Walsh code #32 are used to spread each modulation symbol,

resulting in a x256 rate increase; the resulting 0s and 1s are now called “chips”• there are 32,768 chips in a Sync Channel frame (1024 chips per original bit)

GainControl

BasebandFilter

BasebandFilter

I PN

Q PN

1.2288Mcps

WalshFunction 32

1200 bps BlockInterleaving

R = 1/2, K = 9ConvolutionalEncoding &Repetition

4800 bps 4800 bps

bits

modulationsymbols

chips



Figure 3-33Sync channel block interleaver (input matrix)

The Sync Channel uses a block interleaver spanning 262/3 ms, which is equivalent to 128 modulation symbols) at the symbol rate of 4800 sps. These symbols are arranged in a 16 row by 8 column array.

The input (array write) sequence is shown in the figure above. Symbols are written down by columns, from left to right. Positions with the same number denote repeated symbols.

The arrows show how the read array, appearing in the next page, is built. This technique for interleaving the Sync Channel symbols is sometimes described as a “bit reversal” method.

1 9 17 25 33 41 49 57

1 9 17 25 33 41 49 57

2 10 18 26 34 42 50 58

2 10 18 26 34 42 50 58

3 11 19 27 35 43 51 59

3 11 19 27 35 43 51 59

4 12 20 28 36 44 52 60

4 12 20 28 36 44 52 60

5 13 21 29 37 45 53 61

5 13 21 29 37 45 53 61

6 14 22 30 38 46 54 62

6 14 22 30 38 46 54 62

7 15 23 31 39 47 55 63

7 15 23 31 39 47 55 63

8 16 24 32 40 48 56 64

8 16 24 32 40 48 56 64



Figure 3-34Sync channel block interleaver (output matrix)

The output (array read) sequence is shown in the figure above. Symbols are read down by columns, from left to right.

Assume that during transmission a burst of noise occurs that affects a group of consecutively transmitted symbols.

1 3 2 4 1 3 2 4

33 35 34 36 33 35 34 36

17 19 18 20 17 19 18 20

49 51 50 52 49 51 50 52

9 11 10 12 9 11 10 12

41 43 42 44 41 43 42 44

25 27 26 28 25 27 26 28

57 59 58 60 57 59 58 60

5 7 6 8 5 7 6 8

37 39 38 40 37 39 38 40

21 23 22 24 21 23 22 24

53 55 54 56 53 55 54 56

13 15 14 16 13 15 14 16

45 47 46 48 45 47 46 48

29 31 30 32 29 31 30 32

61 63 62 64 61 63 62 64

assume that a burst of noise affects these symbols



Figure 3-35Sync channel block interleaver (block restored)

The output (array read) sequence is shown in the figure above. Symbols are read down by columns, from left to right.

Notice that when the symbols are restored to their original order, the effect of the noise burst is spread. The possibly corrupted symbols are not consecutive any longer and the error correction mechanisms have a better chance to correct them.

1 9 17 25 33 41 49 57

1 9 17 25 33 41 49 57

2 10 18 26 34 42 50 58

2 10 18 26 34 42 50 58

3 11 19 27 35 43 51 59

3 11 19 27 35 43 51 59

4 12 20 28 36 44 52 60

4 12 20 28 36 44 52 60

5 13 21 29 37 45 53 61

5 13 21 29 37 45 53 61

6 14 22 30 38 46 54 62

6 14 22 30 38 46 54 62

7 15 23 31 39 47 55 63

7 15 23 31 39 47 55 63

8 16 24 32 40 48 56 64

8 16 24 32 40 48 56 64



Figure 3-36Sync channel structure

The Sync Channel message is made of the Message Body (containing the message data), preceded by the 8-bit message length (in bytes) and followed by a 30-bit Cyclic Redundancy Code (CRC).

The Sync Channel Message is contained within the Sync Channel Message Capsule. The length of the capsule is always an exact multiple of 96 bits. (Notice that at the rate of 1200 bps or 1.2 bits per ms, it takes the same exact multiple of 80 ms for the capsule to be transmitted). 96 bits or 80 ms is called a Superframe, and it happens to be exactly three 262/3 ms (32-bit) frames.

If the Sync Channel Message does not fit exactly within the capsule that contains it, enough zeros are padded at the end of the message to match the length of the capsule.

Notice that only 31 bits of each frame (the frame body) are used to carry a portion of the message capsule. The first bit of each frame (SOM) indicates whether this is the Start Of a Message capsule (in which case its value is 1) or not (in which case its value is 0).

26.67 ms32 bits

31 bits

Sync Channel Message Capsule (93 x Ns bits)

Sync Channel Message (8 x MSG_LENGTH)

8 bits 30 bits2-1146 bits

as required

Sync Channel Frame Body

Sync Channel Frame

SOM

80 ms, 96 bits

1200 bps

Ns = Number of SyncChannel Superframes needed for message transmission

Sync Channel Message Padding

MSG_LENGTH Message Body CRC

1 0 0 0 0 0 0

Sync Channel Superframe Sync Channel Superframe



Figure 3-37Sync channel message body format

The Synch Channel Message body is 170-bit long. Add 8 bits for the length and 30 for the CRC, and you have 208 bits.

Divide 208 by 31 (number of bits in a frame that can be used to carry a portion of the message capsule) and you get 6.71 or 6 frames and 22 bits of the seventh frame.

But the capsule is actually made of an integer number of superframes, each one of them taking three frames; therefore, we must round up the number of frames to 9 to accommodate this message. This is three superframes with most of the third superframe filled with padding zeroes.

One superframe takes 80 ms to be transmitted; three take 240 ms. Therefore, the Sync Channel Message is transmitted about four times per second.

MSG_TYPE (‘00000001’)

P_REV

MIN_PREV

SID

NID

PILOT_PN

LC_STATE

SYS_TIME

LP_SEC

LTM_OFF

DAYLT

PRAT

CDMA_FREQ

8

8

8

15

16

9

42

36

8

6

1

2

11

Field Length(bits)

Total : 170



Figure 3-38Sync channel message parameters


Sync Message ParametersSync Message Parameters

• Message Type (MSG_TYPE) – Identifies this message and determinesits structure (set to the fixed value of ‘00000001’)

• Protocol Revision Level (P_REV) – Shall be set to ‘00000001’

• Minimum Protocol Revision Level (MIN_P_REV) – 8-bit unsignedinteger identifying the minimum protocol revision level required to operateon this system. Only personal stations that support revision numbersgreater than or equal to this field can access the system

• System ID (SID) – 16-bit unsigned integer identifying the system

• Network ID (NID) – 16-bit unsigned integer identifying the network withinthe system (defined by the owner of the SID)

• Pilot PN Sequence Offset Index (PILOT_PN) – Set to the pilot PN offsetfor the base station (in units of 64 chips), assigned by the network planner

• Long Code State (LC_STATE) – Provides the mobile station with thebase station long code state at the time given by the SYS_TIME field,generated dynamically

• System Time (SYS_TIME) – GPS system-wide time as 320 ms after theend of the last superframe containing any part of this message, minus thepilot PN offset, in units of 80 ms, generated dynamically



Figure 3-39Sync channel message parameters


Sync Channel Message Parameters (cont)Sync Channel Message Parameters (cont)

• Leap Seconds (LP_SEC) – Number of leap seconds that have occurred sincethe start of system time (January 6, 1980 at 00:00:00 hours) as given in theSYS_TIME field, generated dynamically

• Local Time Offset (LTM_OFF) – Two’s complement offset of local time fromsystem time in units of 30 minutes, generated dynamically

• Current local = SYS_TIME – LP_SEC + LTM_OFF

• Daylight savings time indicator (DAYLT) – Determined by the networkplanner

• 1 if daylight savings in effect in this base station

• 0 otherwise• Paging Channel Data Rate (PRAT) – The data rate of the paging channel for

this system, determined by the network planner

• 00 if 9600 bps

• 01 if 4800 bps• CDMA Frequency Assignment (CDMA_FREQ) – The CDMA channel

number, in the specified CDMA band class, corresponding to the frequencyassignment for the CDMA Channel containing a Primary Paging Channel,determined by the network planner



Paging channels 3The Paging Channels are used by the base station to transmit system overhead information and mobile station-specific messages. For each Paging Channel that the base station transmits, the base station continually transmits valid Paging Channel messages, which may include the “Null Message” which consists of two zero bits.

The Paging Channel transmits information at a fixed data rate of either 9600 or 4800, as specified by the “PRAT” (Paging Channel Data Rate) parameter sent in the Sync Channel Message.

Figure 3-40Paging channels

■ Up to seven paging channels can be supported on a single CDMAfrequency assignment

■ Channel 1 (Walsh function 1) is the Primary Paging Channel■ Additional Paging Channels use Walsh functions 2 through 7■ Unused paging channels can be used as Forward Traffic Channels■ Two rates are supported: 9600 and 4800 bps (PRAT parameter in

the Sync Channel Message)■ A single 9600 bps Paging Channel can support about 180 pages

per second

Used by the base station to transmit system overhead informationand mobile station-specific messages.



Figure 3-41Paging channel generation

The Paging Channel is encoded, interleaved, spread and modulated in 20 ms frames.

In the Paging Channels, the convolutional encoder is not “flushed” with zeroes at the end of each frame. Therefore, the last 8 bits of a Paging Channel frame influence the first 18 (for 4800 bps) or 36 (for 9600 bps) modulation symbols generated from the first 8 bits of the following frame.

The Paging Channel Data is scrambled using the long PN code sequence offset by a 42-bit mask (see next slide).

The Paging Channel is orthogonally spread by the Walsh function with index equal to the Paging Channel number.

The Paging Channel is then “quadrature spread” and “filtered” (discussed later).

■ There are 192 [96] bits (9600 [4800] bps x 0.020 second) in one PagingChannel frame

■ The Rate 1/2 convolutional encoder doubles the bit rate, resulting” 384 [192]code symbols in a Paging Channel frame

■ If the 4800 bps rate is used, the repetition process doubles the rate again, sothat, at either rate, 384 modulation symbols per Paging Channel frame result

■ 384 modulation symbols per frame times 50 frames per second = 19.2 Ksps■ One copy of Walsh code #1 (or #2, ... or #7) is used to spread each modulation

symbol. This results in a x64 rate increase to 1.2288 Mcps• that is, 24,576 chips per Paging Channel frame, or 128 [256] chips per

original bit at 9600 [4800] bps

GainControl

BasebandFilter

BasebandFilter

I PN

Q PN

1.2288Mcps

WalshFunction 1-7

9600 bps4800 bps

BlockInterleaving

R = 1/2, K = 9ConvolutionalEncoding &Repetition

19.2Ksps

19.2Ksps

DecimatorLong PN Code

Generator

Scrambling

PagingChannel

Address Mask

1.2288Msps

19.2Ksps



Figure 3-42Paging channel structure

The Paging Channel is logically divided in 80 ms slots. Each slot is composed of four 20-ms Paging Channel frames, each subdivided into two 10-ms half-frames. The first bit in any half frame is a Synchronized Capsule Indicator (SCI). This bit is set to “1” to indicate that the initial segment of a message capsule immediately follows; otherwise, the SCI bit is set to “0”.

A Paging Channel Message Capsule is composed of a Paging Channel Message and padding. A Paging Channel Message consists of a length field, a message body, and a CRC field. Padding consists of zero or more “0” bits. The message capsule is subdivided into small pieces, each fitting the rest of a half frame (called a Half Frame Body).

The base station may transmit synchronized or unsynchronized Paging Channel message capsules. A synchronized message capsule starts on the second bit of a half frame. An unsynchronized message capsule begins immediately after a previous message capsule.

If, after the end of a Paging Channel message, there remain fewer that 8 bits before the next SCI bit, the base station may transmit an unsynchronized message capsule immediately following that message. The base station will not include any padding bits in a message capsule that is followed by an unsynchronized message capsule.

R = 9600 or 4800 bps

(1) First new capsule in slot, Synchronized Capsule(2) Unsynchronized Capsules(3) Synchronized Capsules

8 x MSG_LENGTHas required

SCI

8 bits 30 bits(see notein text)

163.84 s, 163.84 x R bits

2048 slots

8 Half Frames per Slot

(1) (2) (3)

10 ms

SCI : Synchronized Capsule Indicator

1

Slot Channel 0 Slot Channel ‘n’ Slot Channel 2047

Half Frame Half Frame Half Frame Half Frame Half Frame

Half Frame Body Half Frame Body Half Frame Body Half Frame Body Half Frame Body0 1 0 0 0 1

Message Capsule Message Capsule Messageage Capsule

Paging Channel Message Paging Channel Message PaddingPadding Paging ChMessage

MSG_LENGTH Message Body CRC

Maximum Paging Channel Slot Cycle



Note: Any message sent by the base station on the Paging Channel must be completely contained in one or two consecutive Paging Channel slots.



Exercise 3-1 Lesson Review


1. What vocoder function stores a collection of arbitrary waveform segments?

2. What reference channel is used for acquisition, timing, and as a phase reference for coherent demodulation?

3. Lower data rates are transmitted at reduced power rates. [True / False]

4. The frame duration of what channel matches the period of repetition for of the short PN sequences?

5. The sync channel is identified by what Walsh code function?

6. The pilot channel is identified by what Walsh code function?

7. Convolutional encoding occurs before block interleaving (on the forward channel). [True / False]

8. What is the purpose of the paging channel?

9. What Walsh functions are reserved for the paging channels?

10. Unused paging channels can be used as what type of channel?

11. The effect of bursty errors are minimized by what function?





Lesson 4 CDMA reverse channelsObjectives 4


• identify the CDMA Reverse Traffic Channels

• describe how the Reverse Traffic Channels are generated

• explain the concept of “code channel” in the reverse direction and how this differs from the concept of code channel in the forward direction

4-2 CDMA reverse channels Nortel Networks Confidential


Figure 4-1CDMA code channels in the reverse direction

A Reverse Channel is identified by:

• its CDMA RF carrier frequency

• the unique Long Code PN Offset of the individual handset

REG

1-8002424444

BTS

There are two types of CDMA Reverse Channels:

■ TRAFFIC CHANNELS are used by individualusers during their actual calls to transmit trafficto the BTS• a reverse traffic channel is defined by a user-specific

public or private Long Code mask• there are as many reverse Traffic Channels as

there are CDMA phones in the world

■ ACCESS CHANNELS are used by mobile stationsnot yet in a call to transmit registration requests,call setup requests, page responses, orderresponses, and other signaling information• an access channel is defined by a user-independent

public long code mask• Access channels are paired with Paging Channels.

There can be up to 32 access channels per pagingchannel

Nortel Networks Confidential CDMA reverse channels 4-3


Figure 4-2Coding process on CDMA reverse code channels

• Each mobile is uniquely identified by the offset of the User Long Code, which is generated internally.

• All mobile stations transmit simultaneously on the same 1.25-MHz wide frequency band.

• Any nearby BTS can dedicate a channel element to the mobile station and successfully extract its signal.

• Mobile stations also use the other CDMA spreading sequences, but not for channel-identifying purposes.

• Short PN Sequences are used to achieve phase modulation.

• Walsh Codes are used as “symbols” to give ultra-reliable communications recovery at the BTS.

MTX BSC BTS (1 sector)

Channel Element

Access Channels

Vocoder

Vocoder

Vocoder

Vocoder

more more

Receiver,Sector X

Channel Element

Channel Element

Channel Element

Channel Element

Long Code Gen

Long Code Gen

Long Code Gen

Long Code Gen

Long Code Gen

more

UserLongCode User

LongCode User

LongCode

UserLongCode

UserLongCode

UserLongCode

A Reverse Channel is identified by:❖ its CDMA RF carrier Frequency❖ the unique Long Code PN Offset

of the individual handset

CDMAFrequency



Access channels 4To initiate communication with the base station and to respond to a Paging Channel message, a mobile station uses an Access Channel.

An Access Channel transmission is coded, interleaved, and modulated spread-spectrum signal. The Access Channels use a random-access protocol. Access Channels are uniquely identified by this long codes.

Figure 4-3Access channels

■ Used by the mobile station to• Initiate communication with the base station• Respond to Paging Channel messages

■ Has a fixed data rate of 4800 bps■ Each Access Channel is associated with only one Paging Channel■ Up to 32 access channels (0-31) are supported per Paging Channel

4800 bps



Figure 4-4Access channel generation

Notice the following:

• Access Channels operate at a fixed data rate: 4800 bps

• Access Channels use a rate 1/3 convolutional coder (1 bit in, 3 symbols out)

• A larger block (32 x 18) interleaving array is needed

• The Access Channels cannot use the orthogonal spreading scheme as in the Forward Channels. They use orthogonal modulation and direct sequence spreading instead.

• Both repetitions of each code symbol are transmitted.

• The Q sequence is delayed by half a chip relative to the I sequence

PROBLEM: The mobile station must capture the attention of the base station. The mobile station makes a series of attempts to send a message (Access Probes). Each instance of a message is sent at higher power level than the previous. Messages are sent at random intervals to reduce the probability of collision with messages from other mobile stations trying to reach the same base station on the same Access Channel.

■ Message attempts are randomized to reduce probability of collision■ Two message types:

• A response message (in response to a base station message)• A request message (sent autonomously by the mobile station)

28.8kspsConvolutional

Encoder &Repetition

R = 1/3

1.2288McpsAccess Channel

Long Code MaskLong PN Code

Generator

28.8ksps Orthogonal

Modulation

307.2kcps

1.2288Mcps

Q PN (No Offset)

I PN (No Offset)

D

1/2 PNChipDelay

BlockInterleaver

Access ChannelInformation

(88 bits/Frame)

4.8 kpbs

DirectSequenceSpreading



Figure 4-5Rate 1/3 convolutional encoder

The figure illustrates the Rate 1/3 convolutional coder used by the Reverse Traffic Channels with the 8 kb vocoders.

+

+

+

g0

g1

g2

Information bits(INPUT)

Code Symbols(OUTPUT)



1 2 3 4 5 6 7 8



Figure 4-6Access channel block interleaving

The mobile station uses a 32-row by 18-column array to interleave the 576 modulation symbols coming out of the convolutional coder every 20 ms.

■ 576 code symbols (288 x 2) are written sequentially by columns, thenread by rows in a particular order (called “bit-reverse readout of the rowaddresses”) every 20 ms

■ Block interleaving separates repeated symbols in two identical sets:one set is transmitted during the first 10 ms and the second set, withthe repetitions, is transmitted during the second 10 ms

• Improves survivability of symbol information• “Spreads” the effect of spurious interference and fast fading

28.8 ksps fromConv. Encoding

& SymbolRepetition (2x)

Input Array(Normal

Sequence)32 x 18


32 x 18

28.8 ksps toOrthogonalModulation



Figure 4-7Access channel block interleaving (4800 X 2 bps – WRITE MATRIX)

1 17 33 49 65 81 97 113 129 145 161 177 193 209 225 241 257 273 1 17 33 49 65 81 97 113 129 145 161 177 193 209 225 241 257 273 2 18 34 50 66 82 98 114 130 146 162 178 194 210 226 242 258 274 2 18 34 50 66 82 98 114 130 146 162 178 194 210 226 242 258 274 3 19 35 51 67 83 99 115 131 147 163 179 195 211 227 243 259 275 3 19 35 51 67 83 99 115 131 147 163 179 195 211 227 243 259 275 4 20 36 52 68 84 100 116 132 148 164 180 196 212 228 244 260 276 4 20 36 52 68 84 100 116 132 148 164 180 196 212 228 244 260 276 5 21 37 53 69 85 101 117 133 149 165 181 197 213 229 245 261 277 5 21 37 53 69 85 101 117 133 149 165 181 197 213 229 245 261 277 6 22 38 54 70 86 102 118 134 150 166 182 198 214 230 246 262 278 6 22 38 54 70 86 102 118 134 150 166 182 198 214 230 246 262 278 7 23 39 55 71 87 103 119 135 151 167 183 199 215 231 247 263 279 7 23 39 55 71 87 103 119 135 151 167 183 199 215 231 247 263 279 8 24 40 56 72 88 104 120 136 152 168 184 200 216 232 248 264 280 8 24 40 56 72 88 104 120 136 152 168 184 200 216 232 248 264 280 9 25 41 57 73 89 105 121 137 153 169 185 201 217 233 249 265 281 9 25 41 57 73 89 105 121 137 153 169 185 201 217 233 249 265 28110 26 42 58 74 90 106 122 138 154 170 186 202 218 234 250 266 28210 26 42 58 74 90 106 122 138 154 170 186 202 218 234 250 266 28211 27 43 59 75 91 107 123 139 155 171 187 203 219 235 251 267 28311 27 43 59 75 91 107 123 139 155 171 187 203 219 235 251 267 28312 28 44 60 76 92 108 124 140 156 172 188 204 220 236 252 268 28412 28 44 60 76 92 108 124 140 156 172 188 204 220 236 252 268 28413 29 45 61 77 93 109 125 141 157 173 189 205 221 237 253 269 28513 29 45 61 77 93 109 125 141 157 173 189 205 221 237 253 269 28514 30 46 62 78 94 110 126 142 158 174 190 206 222 238 254 270 28614 30 46 62 78 94 110 126 142 158 174 190 206 222 238 254 270 28615 31 47 63 79 95 111 127 143 159 175 191 207 223 239 255 271 28715 31 47 63 79 95 111 127 143 159 175 191 207 223 239 255 271 28716 32 48 64 80 96 112 128 144 160 176 192 208 224 240 256 272 28816 32 48 64 80 96 112 128 144 160 176 192 208 224 240 256 272 288



Figure 4-8Access channel block interleaving (4800 X 2 bps – READ MATRIX)

Notice that after rearranging the symbols we end up with two identical sets.

Both sets are transmitted at full power, one during the first 10 ms and the other during the second 10 ms.

1 17 33 49 65 81 97 113 129 145 161 177 193 209 225 241 257 273 9 25 41 57 73 89 105 121 137 153 169 185 201 217 233 249 265 281 5 21 37 53 69 85 101 117 133 149 165 181 197 213 229 245 261 27713 29 45 61 77 93 109 125 141 157 173 189 205 221 237 253 269 285 3 19 35 51 67 83 99 115 131 147 163 179 195 211 227 243 259 27511 27 43 59 75 91 107 123 139 155 171 187 203 219 235 251 267 283 7 23 39 55 71 87 103 119 135 151 167 183 199 215 231 247 263 27915 31 47 63 79 95 111 127 143 159 175 191 207 223 239 255 271 287 2 18 34 50 66 82 98 114 130 146 162 178 194 210 226 242 258 27410 26 42 58 74 90 106 122 138 154 170 186 202 218 234 250 266 282 6 22 38 54 70 86 102 118 134 150 166 182 198 214 230 246 262 27814 30 46 62 78 94 110 126 142 158 174 190 206 222 238 254 270 286 4 20 36 52 68 84 100 116 132 148 164 180 196 212 228 244 260 27612 28 44 60 76 92 108 124 140 156 172 188 204 220 236 252 268 284 8 24 40 56 72 88 104 120 136 152 168 184 200 216 232 248 264 28016 32 48 64 80 96 112 128 144 160 176 192 208 224 240 256 272 288

1 17 33 49 65 81 97 113 129 145 161 177 193 209 225 241 257 273 9 25 41 57 73 89 105 121 137 153 169 185 201 217 233 249 265 281 5 21 37 53 69 85 101 117 133 149 165 181 197 213 229 245 261 27713 29 45 61 77 93 109 125 141 157 173 189 205 221 237 253 269 285 3 19 35 51 67 83 99 115 131 147 163 179 195 211 227 243 259 27511 27 43 59 75 91 107 123 139 155 171 187 203 219 235 251 267 283 7 23 39 55 71 87 103 119 135 151 167 183 199 215 231 247 263 27915 31 47 63 79 95 111 127 143 159 175 191 207 223 239 255 271 287 2 18 34 50 66 82 98 114 130 146 162 178 194 210 226 242 258 27410 26 42 58 74 90 106 122 138 154 170 186 202 218 234 250 266 282 6 22 38 54 70 86 102 118 134 150 166 182 198 214 230 246 262 27814 30 46 62 78 94 110 126 142 158 174 190 206 222 238 254 270 286 4 20 36 52 68 84 100 116 132 148 164 180 196 212 228 244 260 27612 28 44 60 76 92 108 124 140 156 172 188 204 220 236 252 268 284 8 24 40 56 72 88 104 120 136 152 168 184 200 216 232 248 264 28016 32 48 64 80 96 112 128 144 160 176 192 208 224 240 256 272 288



Figure 4-9Access channel slot structure

ACCESS CHANNEL SLOT – Group of contiguous Access Channel frames where an Access Channel Message can be contained. As the Access Slot length may differ from base station to base station, a mobile station must determine the beginning and length of the Access Channel slot prior to transmission.

All access channels associated with a particular Paging Channel have the same slot size, and all the slots begin at the same time.

Each Access Channel Frame contains 96 bits (20 ms frame at 4800 bps). Each Access Channel frame consists of 88 information bits and eight Encoder Tail Bits.

The Encoder Tail Bits are a fixed sequence of bits (eight zeroes) added to the end of a block of data (88 bits) to reset (flush) the convolutional encoder to a known state (all zeroes).

The Access Channel Preamble consists of frames of 96 zeroes that are transmitted at the 4800 bps rate. The Access Channel Preamble is transmitted to aid the base station in acquiring an Access Channel transmission.

Access Channel Message Padding

CRC

20 ms96 bits

8 x MSG_LENGTH

8 bits 30 bits

MSG_LENGTH

2-842 bits

asrequired

Access Channel Frame

Message Body

20 x (4 + PAM_SZ + MAX_CAP_SZ) ms96 x (4 + PAM_SZ + MAX_CAP_SZ) bits

4800 bps

Nf =Number of AccessChannel Frames

needed for message transmissionT = Encoder Tail Bits

(eight zeroes)

Access Channel Frame Body

1 + PAM_SZ frames96 x (1 + PAM_SZ) bits

96 x Nf bits (not exceeding 3 + MAX_CAP_SZ frames)

Access Channel Slot

TAccess Channel Preamble

Access Channel Message Capsule

T TT

88 x Nf bits



Figure 4-10Access channel probing

AccessProbe 1

AccessProbe 1

AccessProbe 1

AccessProbe 1

Access Probe1 + NUM_STEP

(16 max)

SystemTime

TA RT TA RT TA RT TA

PI

PI

PI

IP(Initial

Power)

See previousfigure

ACCESSPROBE

SEQUENCE

Select Access Channel (RA)initialize transmit power

AccessProbe 1

AccessProbe 1

AccessProbe 1

AccessProbe 1

Access Probe1 + NUM_STEP

(16 max)

SystemTime

TA RT TA RT TA RT TA

PI

PI

PI

IP(Initial

Power)

See previousfigure

ACCESSPROBE

SEQUENCE

Select Access Channel (RA)initialize transmit power



Reverse traffic channels 4A mobile station using the 8 kb vocoder transmits information on the Reverse Traffic Channel at variable data rates of 9600, 4800, 2400, and 1200 bps (Rate Set 1, Multiplex Option 1) as in the Forward Traffic Channels.

A mobile station using the 13 kb vocoder transmits information on the Reverse Traffic Channel at variable data rates of 14400, 7200, 3600, and 1800 bps (Rate Set 2, Multiplex Option 2) as in the Forward Traffic Channels.

The frame duration is 20 ms.

Figure 4-11CDMA reverse traffic channels

■ Used when a call is in progress to send• Voice traffic from the subscriber• Response to commands/queries from the base station• Requests to the base station

■ Supports variable data rate operation for• 8 Kbps vocoder

− Rate Set 1 - 9600, 4800, 2400 and 1200 bps− Multiplex Option 1

• 13 Kbps vocoder− Rate Set 2 - 14400, 7200, 3600, 1800 bps− Multiplex Option 2



Figure 4-12Reverse traffic channel generation

Notice the differences:

• A Rate 1/3 (1 bit in, 3 symbols out) convolutional encoder is used with 8Kb vocoders/Rate Set 1, while a Rate 1/2 (1 bit in, 2 symbols out) convolutional encoder without symbol puncturing is used with 13Kb vocoders/Rate Set 2.

• More symbols in and out of the block interleaver each 20 ms; therefore, a larger matrix is needed. Different array reading discipline, depending on the frame rate.

• Can’t use the orthogonal spreading scheme as in the Forward Channels (each reverse channel comes from a different mobile station). Uses Orthogonal modulation instead.

• The “Data Burst Randomizer” follows the Orthogonal Modulation step.

• A 1/2 chip delay is introduced in the Q sequence.

• The I and Q pilot PN sequences are not offset relative to the mobile station time. However they are offset by an undetermined number of chips relative to the system time because the mobile draws its time reference from the earliest usable pilot component that it receives (which arrives with a delay that depend on the distance between the mobile station and the base station where that pilot signal was generated).

• Direct Sequence Spreading involving every Long PN sequence chip instead of Data Scrambling involving one out of every 64 is used to differentiate among mobile stations.

Notice that by the time the first frame arrives on the Forward Traffic Channel, the mobile station has had plenty of time to acquire the CDMA system (first the pilot channel, then the synch channel, then the paging channel). This is

9600 bps4800 bps2400 bps1200 bps

or 14400 bps7200 bps3600 bps1800 bps

28.8ksps

R = 1/3

1.2288McpsUser Address

Mask

LongPN Code

Generator

28.8ksps Orthogonal

ModulationData Burst

Randomizer

307.2kcps

1.2288Mcps

Q PN(no offset)

I PN(no offset)

D

1/2 PNChipDelay

DirectSequenceSpreading

R = 1/2

ConvolutionalEncoder &Repetition

BlockInterleaver



not the case in the reverse direction. The mobile station must therefore give the base station a chance to acquire the Reverse Traffic Channel before it starts sending information. That is why the first few frames sent by the mobile station to the base station do not contain information but just zeroes. This is called the Reverse Traffic Channel Preamble.

The Reverse Traffic Channel Preamble is the first frame transmitted in a Reverse Traffic Channel. It consists of all zeroes transmitted with a 100% transmission duty cycle (192 zeroes for Rate 1 and 288 for Rate 2).

The Reverse Traffic Channel preamble is transmitted by the mobile station to help the base station acquire the Reverse Traffic Channel.

Figure 4-13Reverse traffic channel frame structure

Each time a mobile station receives a bad Rate Set 2 Forward Traffic Channel frame, it sets the “erasure bit to ‘1’ in the second Rate Set 2 Reverse Traffic Channel frame transmitted following the reception of the bad frame. Otherwise it sets the erasure bit to ‘0’.

TransmissionRate Total Erasure Information CRC Tail Bits

9600 192 — 172 12 8

4800 96 — 80 8 8

2400 48 — 40 — 8

1200 24 — 16 — 8

14400 288 1 267 12 8

7200 144 1 125 10 8

3600 72 1 55 8 8

1800 36 1 21 6 8

1

2

Number of Bits per FrameRateSet



Figure 4-14Reverse traffic channel: Convolutional encoding and symbol repetition

When the 8 kb vocoders are used, the Reverse Traffic Channels employ a Rate 1/3, K=9 convolutional coder which triplicates the 9.6 kbps into 28.8 ksps. When the 13 kb vocoders are used, the Reverse Traffic Channels employ a Rate 1/2, K=9 convolutional coder which duplicate the 14.46 kbps into 28.8 ksps. Symbol repetition ensures that a constant output of 28.8 ksps is fed to the block interleaver when any rate other than a full rate is used.

■ Convolutional encoding:• Results in 3 code symbols out for each bit in, at Rate Set

1, and in 2 code symbols out for each bit in, at Rate Set 2• Also allows for reduction in transmit power• Reduces overall noise & increases capacity

■ Symbol repetition maintains a constant 28.8 ksps output toblock interleaver

PCM Voice



BlockInterleavingOrthogonalModulationData Burst

RandomizerDirect Sequence

SpreadingQuadratureSpreadingBasebandFiltering

VocoderProcessing


VariableRate

Outputfrom

Vocoder

R=1/3 K=9 Convolutional

EncoderR=1/2 K=9

28.8 kspsto Block

Interleaver

28.8 ksps (No repetition)14.4 ksps (2 X repetition)7.2 ksps (4 X repetition)3.6 ksps (8 X repetition)

28.8 ksps (No repetition)14.4 ksps (2 X repetition)7.2 ksps (4 X repetition)3.6 ksps (8 X repetition)

SymbolRepetition

9.6 kbps4.8 kbps2.4 kbps1.2 kbps

14.4 kbps 7.2 kbps 3.6 kbps 1.8 kbps



Figure 4-15Reverse traffic channel: Block interleaving

The mobile station uses a 32-row by 18-column array to interleave the 576 modulation symbols coming out of the convolutional coder every 20 ms.

■ 20 ms symbol blocks are sequentially reordered■ Combats the effects of fast fading■ Separates repeated symbols at 4800 bps and

below• Improves survivability of symbol data• “Spreads” the effect of spurious interference

PCM Voice



BlockInterleaving

VocoderProcessing


28.8 kspsFrom Coding& SymbolRepetition


32 x 18

28.8 ksps toOrthogonalModulation

OrthogonalModulationData Burst

RandomizerDirect Sequence

SpreadingQuadratureSpreadingBasebandFiltering

Input Array(Normal

Sequence)32 x 18





1. The two types of CDMA Reverse Channels are Traffic Channels and Access Channels. [True/False]

2. Short PN sequences are used to achieve ________________________ .

3. How many access channels are supported by a single paging channel?

4. When generating the Access Channel, why are message attempts randomized?

5. All Access Channels associated with a particular Paging Channel

a. have the same slot size

b. do not have the same slot size

c. all slots begin at the same time

d. all slots do not begin at the same time

e. a and c

f. a and d

g. none of the above

6. What is the access channel preamble?

7. Why is the Reverse Traffic Channel preamble transmitted by the mobile to the base station?

8. The pilot PN sequences are offset relative to system time, not mobile station time. Why?



9. What is used in the reverse path: direct sequence spreading or data scrambling?



Lesson 5 Power control, registration, and handoffsObjectives 5


• understand the purpose of Power Control in CDMA

• identify the different types of Power Control mechanisms used in CDMA

• define the term registration

• recall the differences between “HLR” and “VLR”

• define the concept of “handoff” and identify its three phases

• identify the different cases of CDMA handoffs

5-2 Power control, registration, and handoffs Nortel Networks Confidential


CDMA power control 5All code channels transmitted in the forward direction travel together to the mobile station; therefore, they all experience fading at the same time. This is not the case in the reverse direction, where each channel travels a different path from the mobile station that transmits it to the base station. That is the reason why the use of a pilot signal and “coherent demodulation” is possible in the forward direction and not in the reverse direction, where the mobile stations have to resort to the transmission of “preambles” of one sort or another.

Some mobile stations may be close to the base station, while others may be located far from it. As a result, the path losses and multipath environments affecting the signals from different mobile stations show a great variability. As the path losses can differ by up to 80 dB, if every mobile station transmits at the same power level, the base station could receive a very strong signal from a nearby mobile station, together with another signal, 80 times weaker, from a distant one; and the weaker signal would be drowned by the stronger one.

The purpose of Power Control is to ensure that all signals arrive at the base station at approximately the same level. This requirement makes power control in the reverse direction extremely critical and demanding.

Nortel Networks Confidential Power control, registration, and handoffs 5-3


Figure 5-1CDMA power control

The goal of power control is that the signal of interest is received with sufficient strength so that the demodulator has enough Eb/N0 to recover the signal with an acceptable level of errors; while the remaining signals, which from the point of view of this mobile station, are just interference, and are kept at a power level that is as low as possible.

As what one receiver perceives as interference happens to be the signal of interest for another (and vice versa), everyone must be kept at the “near starvation” level for the general good.

The first step taken by a mobile station to control its transmit power is called Reverse Open Loop Power control. It essentially consists of estimating how strong the mobile station should transmit based on a coarse measurement of how much power it is receiving from the base station and some correcting parameters delivered in the Access Parameters Message.

■ CDMA is an interference-limited system based on the number of users

■ Unlike AMPS/TDMA, CDMA has a soft capacity limit

• Each user is a noise source on the shared channel

• The noise contributed by users is cumulative

• This creates a practical limit to how many users a system will sustain

■ Precise power control of the mobile stations is critical if we want to

• Maximize system capacity

• Increase battery life of the mobile stations

■ The goal is to keep each mobile station at the absolute minimumpower level necessary to ensure acceptable service quality

• Ideally the power received at the base station from each mobile stationshould be the same (minimum signal to interference)

• Mobile stations which transmit excessive power increase interference toother mobile stations



Figure 5-2Reverse open loop power control

• The mobile station makes a coarse initial estimation of the required transmitpower, based upon the total received power.

• Problems with Reverse Open Loop Power Control:

• Assumes same exact path loss in both directions; therefore, cannotaccount for asymmetrical path loss

• Estimates are based on total power received; therefore the power receivedfrom other cell sites by mobile station introduces inaccuracies

• The mobile station makes a coarse initial estimation of the required transmitpower, based upon the total received power.

• Problems with Reverse Open Loop Power Control:

• Assumes same exact path loss in both directions; therefore, cannotaccount for asymmetrical path loss

• Estimates are based on total power received; therefore the power receivedfrom other cell sites by mobile station introduces inaccuracies

BTSMobile

Reverse Open LoopPower Control

BTS

BTS



Figure 5-3Estimated reverse open loop output power

The very first attempt by the mobile station to talk to the base station is designed to probably fail. This is determined by the value of the INIT_PWR parameter supplied by the base station.

mean output power (dBm) = - mean power input (dBm)+ K+ NOM_PWR - 16 x NOM_PWR_EXT+ INIT_PWR

Power output level for the initial probe during open loop probingon the Access Channel (with closed loop correction inactive):

Subsequent probes in the sequence are sent at increased power levels(each probe is incremented by a value equal to the parameter PWR_STEP)

The “turn around constant” K is calculated assuminga nominal cell Effective Radiation Power (ERP) of 5 W

and a nominal cell loading of 50%.

Its value is -73 for cellular systems and -76 for PCS systems

Power output level for the initial transmission on the Reverse Traffic Channel:Power output level for the initial transmission on the Reverse Traffic Channel:

mean output power (dBm) = - mean power input (dBm)+ K+ NOM_PWR - 16 x NOM_PWR_EXT+ INIT_PWR+ the sum of all access probe corrections (dB)



Figure 5-4Reverse closed loop power control

The Reverse Closed Loop Power Control mechanism provides a correction on the Reverse Traffic Channel mean output power level with respect to the Open Loop estimate. 800 times per second (or once every 1.25 milliseconds) the base station overwrites one (13 kb vocoder) or two (8 kb vocoder) code symbols with a “power up” or “power down” command based on the strength of the signal received from the mobile during the preceding 1.25 ms interval. These power control bits are always transmitted at full power.

As during some 1.25 ms intervals the mobile station’s transmitter is “gated on”, and during some other 1.25 ms intervals it is “gated off”, not every “power up” command received from the base station is meaningful. A power control bit is considered valid (and acted upon) only if it is received in the second 1.25 interval following an interval during which the mobile station transmitted. The mobile station “locks” on the accumulation of valid level changes and ignores received power control bits related to gated-off periods when the transmitter is disabled.

Following the reception of a valid power control bit, the mean output power of the mobile station must be within ±0.3 dB of the final value in less than

• Compensates for asymmetries between the forward and reverse paths• Consists of power up (0) and power down (1) commands sent to the mobile

stations, based upon their signal strength, measured at the Base Station andcompared to a specified threshold (setpoint)

• Each command requests a 1dB increment or decrement of the mobilestation transmit power

• Transmitted 800 times per second, always at full power• Allows to compensate for the effects of fast fading

Mobile BTS

Signal StrengthMeasurement

Setpoint

or

Reverse Closed LoopPower Control



500 µs. The change in mean output power level per single valid power control bit is ±1 dB nominal, and the total change in mean output power is the accumulation of all these individual changes. The actual change must be within ±0.5 dB of the nominal value, and the change in mean output power level per 10 valid power control bits of the same sign must be within ±20% of 10 times the nominal change.

Once the first power control bit has been received, after initializing the Reverse Traffic Channel transmissions, the mean output power is defined by the formula in the illustration above.

Following a step change in mean input power, ýPin, the mean output power of the mobile station will experience a corresponding transition in the direction opposite in sign to ýPin.

Figure 5-5Power output estimations (summary)

■ 6JG OGCP KPRWV RQYGT KU FGHKPGF CU VJG RQYGT TGEGKXGF CV VJG OQDKNGUVCVKQP HQT VJG �� /*\ 4( EJCPPGN DCPFYKFVJ DGKPI WUGF

■ 6JG CEEGUU RCTCOGVGT XCNWGU ECP XCT[ DGVYGGP DCUG UVCVKQPU

• 6JG[ CEEQOOQFCVG FKHHGTKPI HQTYCTF RQYGT NGXGNU CPFCPVGPPC ICKPU

• 6JG[ CTG URGEKHKGF KP VJG #EEGUU 2CTCOGVGTU /GUUCIG�UGPV QXGT VJG 2CIKPI EJCPPGN

1RGP .QQR 2TQDKPI QP

VJG #EEGUU %JCPPGN

4GXGTUG 6TCHHKE %JCPPGN

+PKVKCN VTCPUOKUUKQP

#HVGT (KTUV 2QYGT %QPVTQN

$KV *CU $GGP 4GEGKXGF

$CUGF QP�

�� OGCP KPRWV RQYGTOGCUWTGF CV OQDKNG UVCVKQP

�� CEEGUU RCTCOGVGTURTQXKFGF D[ VJG DCUG UVCVKQP

5WDUGSWGPV RTQDGU CTGCFLWUVGF D[ CEEGUU

RCTCOGVGTU RTQXKFGF D[VJG DCUG UVCVKQP�

$CUGF QP�



�� UWO QH CNN VJG CEEGUU

RTQDG CFLWUVOGPVU

$CUGF QP�



�� UWO QH CNN VJG CEEGUU

RTQDG CFLWUVOGPVU�� UWO QH CNN VJG RQYGT

EQPVTQN DKVU TGEGKXGF



Figure 5-6Reverse outer loop power control

If the received power from the mobile station, as measured at the base station, is below the specified threshold (setpoint), the base station sends a ‘0’ power control bit directing the mobile station to raise its output power. If it is higher, the base station sends a ‘1’ power control bit directing the mobile station to lower its output power.

The setpoint itself is raised or lowered by the Reverse Outer Loop Power Control in order to target the desired Frame Error Rate (FER) level, which is typically 1 percent.

• Not part of IS-95A or J-STD-008.• Most gradual form of reverse link error control

• Setpoint is varied according to the FER on the Reverse TrafficChannel (determined at the Base Station Controller)

• Sampled at a rate of 50 frames per second (20 ms / frame)• Setpoint adjusted every 1-2 seconds

FER

Mobile BTS BSC

Reverse Outer Loop Power

Control


Setpoint

or




Figure 5-7Forward traffic channel power control

To support Forward Traffic Channel power control, the mobile station reports FER statistics to the base station. If the base station enables periodic reporting, the mobile station reports FER statistics at specified intervals. If the base station enables threshold reporting, the mobile station reports FER statistics when the frame error rate exceeds a specified threshold.

Either, or both types of reporting can be enabled or disabled at any given time by the base station. Periodic reporting is controlled by PWR_PERIOD_ENABLE and PWR_REP_FRAMES. Threshold reporting is controlled by PWR_THRESH_ENABLE and PWR_REP_THRESH.

The mobile station maintains a counter of the total number of received frames and a counter for the number of received bad frames. At the end of the specified period or when the threshold is exceeded, depending on what has been enabled, the mobile station sends a “Power Measurement Report Message” to the base station. Then it resets both counters to zero and freezes them for PWR_REP_DELAY x ZERO (0) frames following the first transmission of the message. These five parameters are delivered to the mobile station in the System Parameters Message.

• The base station slowly decreases power to each mobile station• As the FER (determined at the mobile station) increases, the mobile

station requests a Forward Traffic Channel power increase

FER

Mobile BTS BSC

Adjust Fwd.power

Forward Link Power Control



Figure 5-8Summary of all power control mechanisms

• All types of power control work together to minimizes powerconsumption at the mobile stations, and increases the overall capacityof the system transmit power

FER FER

Mobile BTS BSC


Setpoint

or

Adjust Fwd.power

Reverse Outer Loop Power

Control


Forward Link Power Control

Reverse Open LoopPower Control



Registration 5A CDMA system could offer to their subscribers the options of local and extended service. This would encompass the cities of Dallas and Fort Worth.

Subscribers with the local service option would either be at their home location when in Dallas and NID roaming when accessing the service in Fort Worth, or at their home location when in Fort Worth and NID roaming when accessing the service in Dallas.

Subscribers with the extended service option would be considered at their home location (not roaming) both in Dallas and in Fort Worth.

Figure 5-9Roaming

■ A mobile station may be in any of the following roaming states:• Home: mobile station is at its home location (not roaming)• NID roaming: mobile station is within a foreign NID but in the home SID• SID roaming: mobile station is within a foreign SID

■ A mobile station maintains a list of one or more “home pairs”• These are SID/NID combinations defining the mobile station’s home

location• They are stored in semi-permanent memory

■ The identity of current SID/NID is contained in the System ParametersMessage (sent on the Paging Channel)

HomeSID/NID List

(2, 3)(2, 0)(3, 1)

SID = 2SID = 4

SID Roaming

NID = 7 NID =3NID = 0

NID Roaming Not RoamingRoaming Status



Figure 5-10HLR and VLR

■ Contains permanent subscriber data• provisioning information• service information• features available to the

subscriber■ Contains dynamic information

• mobile station’s current location■ Supports call routing■ Queried by the MTX when

subscriber information is needed,regardless of the mobile station’scurrent location

■ Stores a subset of the HLRinformation pertaining to the mobilestations currently registered in theVLR’s service area

HLR VLR



Figure 5-11CDMA registration

Registration is the process by which a mobile station notifies the base station of its location, status, identification, slot cycle, and other characteristics. The base station makes use of location information to efficiently page the mobile station when establishing a mobile station-terminated call.

Registration also provides the mobile station’s SLOT_CYCLE_INDEX and SLOTTED_MODE parameters, so that the base station can determine what Paging Channel slots a mobile station operating in the slotted mode is monitoring; and the mobile station’s protocol revision number, so that the base station knows the capabilities of the mobile station.

■ Registration is the means by which a mobile station notifies the cellularsystem of its location, status, identification, and other characteristics

■ Balance is required between paging and registration

• Infrequent registration results in a high rate of paging

• Frequent registration places a high load on access channels

■ Proper system design allows a base station to efficiently page the mobilestation when establishing a mobile-terminated call

■ Registration also provides

• The mobile station’s SLOT_CYCLE_INDEX and SLOTTED_MODE

• The mobile station class mark and protocol revision number so thatthe base station will know the mobile station’s capabilities

■ Two types of mobile registration

• Non-Autonomous: explicitly requested by the base station, or impliedbased on other types of messages received by the mobile station

• Autonomous: triggered by some event other than the reception of anexplicit or implicit request from the base station



Figure 5-12Forms of CDMA registration

CDMA supports nine forms of registration (discussed in the following slides). These forms of registration fall into two groups: “autonomous registration” and “non-autonomous registration.”

The five forms in the autonomous registration group are conditioned, in part, by the roaming status of the mobile station, and by indicators contained in the System Parameters Message.

The four forms in the non-autonomous registration group are triggered by the base station, either explicitly, by sending a Mobile Registration Order to the mobile station; or implicitly, based on the reception at the mobile station of some other types of messages.

■ Power-up registration

■ Power-down registration

■ Timer-based registration

■ Zone-based registration

■ Distance-based registration

■ Parameter-change registration

■ Implicit registration

■ Ordered registration

■ Traffic channel registration

AutonomousRegistration

AutonomousRegistration

Registration Types NOTSupported by Nortel

Registration Types NOTSupported by Nortel

Non-AutonomousRegistration

Non-AutonomousRegistration

All types of registration can be enabled or disabled by means of the System Parameters Message

All types of registration can be enabled or disabled by means of the System Parameters Message



Figure 5-13Power-up registration

Power-up registration is performed when the mobile station is turned on. To prevent multiple registrations when power is quickly turned on and off, the mobile station delays T57m = 20 seconds before registering after entering the Mobile Station Idle State.

The mobile station maintains a power-up / initialization timer. While this timer is active, the mobile station does not attempt to make registration access attempts.

■ Mobile station registers when• Directed to power-on by the user• Switched to an alternate serving system• Switched from using an analog system

■ Delays 20 seconds• Preventing multiple registrations whenever power is quickly

turned on and off

ON�Access Channel



Figure 5-14Power-down registration

Power-down registration is performed when the user directs the mobile station to power off. If power-down registration is performed, the mobile station does not power down until after completing the registration attempt.

The mobile station does not perform power-down registration if it has not previously registered in the system that corresponds to the current SIDs and NIDs.

■ Mobile station registers when directed to power-down by the user■ Mobile station will not power down until attempt is completed■ Mobile station will not do power down registration if

• Not registered in the current system■ Prevents unnecessary attempts to reach a user

• Can be unreliable (v.gr., user powers down in garage)

OFF�Access Channel



Figure 5-15Timer-based registration

Timer-based registration causes the mobile station to register at regular intervals. Its use also allows the system to automatically de-register mobile stations that did not perform a successful power-down registration. Timer-based registration uses a Paging Channel slot counter (equivalent to a timer with increments of 80 ms). Timer-base registration is performed when the counter reaches a maximum value of REG_COUNT_MAX that is controlled by the base station via the REG_PDR field in the System Parameters Message. The base station disables timer-based registration by setting REG_PRD to zero.

The timer expiration time is calculated as REG_COUNT_MAX = int (2 REG_PRD / 4).

The counter is reset on power-up and when switching between different PCS frequency blocks, different band classes, or alternate operating modes. The counter is also reset after each successful registration.

■ Mobile station registers when a timer expires■ Registration period is determined by the base station■ Allows system to de-register mobile stations that fail to register on

power-down

�Access Channel



Figure 5-16Distance-based registration

Distance-based registration causes a mobile station to register when the distance between the current base station and the base station in which it last registered exceeds the threshold defined by the parameter REG_DIST sent to the mobile station in the System Parameters Message. The base station may disable distance-based registration by setting REG_DIST to zero.

The mobile station determines that it has moved a certain distance based on the difference in latitude and longitude between the current base station and the base station where it registered last. The mobile station calculates this distance using the following formula:

where ∆lat = BASE_LAT - BASE_LAT_REG∆long = (BASE_LONG - BASE_LONG_REG) cos (¼ / 180 x BASE_LAT_REG / 14400)

■ Mobile Station MS registers whenever it does an “Idle Handoff” (handoff when not in a call) into a cell which lays outside a circle with REG_DIST radius and centered at the base station where MS last registered• At position “a” MS registers with Base Station BS-1. BS-1 transmits its latitude and

longitude, and the REG_DIST parameter on its paging channel• At position “b” MS does an idle handoff into BS-2 and reads the latitude and longitude

of this base station. MS then calculates the distance between BS-2 and BS-1, and if the result is less than REG_DIST it does not have to re-register

• At position “c” MS is still listening to BS-2 (no need to re-register yet)• At position “d” MS does an idle handoff into BS-3. MS reads the latitude and

longitude of BS-3 and calculates the distance between BS-3 and BS-1. As this distance exceeds REG_DIST, MS re-registers

a

b

cd

BS-1

BS-2

BS-3

REG_DIST� Paging Channel� Access Channel

Idle Handoff

Idle Handoff

a

b

cd

BS-1

BS-2

BS-3

REG_DIST� Paging Channel� Access Channel

Idle Handoff

Idle Handoff

REG D– IST int ∆lat( )2 ∆long( )2+16

---------------------------------------------------

=



Figure 5-17Zone-based registration

Registration zones are groups of base stations within a given system and network. A base station’s zone is identified by the REG_ZONE field of the System Parameters Message.

Zone-based registration causes a mobile station to register whenever it moves into a new zone not on its internally stored list of visited registration zones. A zone is added to the list whenever a registration (including implicit registration) occurs, and is deleted upon expiration of a timer. After a system access, timers are enabled for every zone except the one where the registration attempt was successful. Timers are also enabled at the start of a call.

A mobile station can be registered in more than one zone. Zones are uniquely identified by a zone number (REG_ZONE) plus the SID and NID of the zone.

The mobile station maintains a list of the zones where it has registered. This list must have at least N9m = 7 entries, each including the zone number, its SID and its NID. The exact number of entries in which a mobile station may be considered registered is determined by the TOTAL_ZONES parameter sent by the base station in the System Parameters Message.

CDM A Technology O verv iew February , 2000 - P age 4 -16

Zone-B ased R egistration

■ The m obile station reg isters w hen it enters a new zone■ A zone is a subse t of the base stations w ith in a network■ The m obile station keeps a lis t o f the zones where it has registe red , up to a

m axim um determ ined by the base sta tion■ Each zone is un iquely identified by the registra tion zone num ber param eter

(RE G _ZO N E) plus the S ID and the N ID to w hich it belongs■ The m obile station activates a tim er fo r every zone w here it has reg istered,

except the ac tive one, and de-registe rs when the tim er exp ires■ The m obile station w ill not re -reg ister if it en te rs a zone wh ich is a lready in its lis t

1 2

3

4 5

N O TE : These a rereg is tra tion zones,not TM SI zones!



The mobile station also maintains a timer for each entry in the list. When an entry is removed from the list, the corresponding timer is disabled. The duration of the timer is determined by ZONE_TIMER parameter in the System Parameters Message.

Figure 5-18Parameter-change registration

Parameter-change registration is performed when a mobile station modifies any of the stored parameters or the supported capabilities indicated in the above illustration.

Parameter-change registration is independent of the roaming status of the mobile station.

■ The mobile station registers after it modifies any of the followingparameters (stored in the mobile station):

• the preferred slot cycle index• the slotted mode indicator• the call termination enabled indicators

■ or the following capabilities supported by the mobile station:• the band classes• the power classes• the rates• the operating modes

�Access Channel

SLOT_CYCLE_INDEXSLOTTED_MODE

MOB_TERM_HOMEetc.

SLOT_CYCLE_INDEXSLOTTED_MODE

MOB_TERM_HOMEetc.



Figure 5-19Implicit registration

Whenever an Origination Message or a Page Response Message is sent, the base station can infer the location of a mobile station. This is considered an implicit registration.

■ Occurs when the mobile station and base station exchangemessages not directly related to registration

• Messaging conveys sufficient information to identify mobilestation and its location

■ Considered successful whenever mobile station sends anOrigination Message or Page Response Message

■ Compatible with AMPS and IS-54 methods■ Effectiveness considered adequate to preclude use of ordered

registration

�Access Channel

Origination Message



Figure 5-20Ordered and traffic channel registration

Ordered registration allows the base station to order a mobile station to register by sending a Registration Request Order to the mobile station over the Paging Channel. The mobile station responds to this order with a Registration Message. This type of registration IS NOT supported by Nortel CDMA. There is limited need in the current DMS-MTX implementation for this capability, except perhaps as part of the algorithm for removing entries from the VLR. However, the overhead incurred in attempting to successful order the mobile station to register prohibits using this type of registration effectively.

Traffic Channel registration allows the base station to obtain registration information about a mobile station that has been assigned to a Traffic Channel by sending a Status Request Order message. The mobile station responds with a Status Message and the base station affirms that the mobile is registered by issuing a Mobile Station Registered Message which supplies the mobile station with the current system parameters for the serving base station. This type of registration IS NOT supported by Nortel CDMA. There is no recognized need in the current DMS-MTX implementation for this capability

■ Ordered Registration• Allows the base station to order a mobile station to register

− mobile station can be idle or in an active call■ Traffic Channel Registration

• Allows the base station to obtain registration information about amobile station that has been assigned to a Traffic Channel

• Information exchange occurs on the Traffic Channel• Suggested use is on inter-system handoffs

■ Neither one is supported by Nortel’s CDMA system

� Paging Channel� Access Channel� Forward Traffic Channel� Reverse Traffic Channel

Registration Message

Traffic Channel Registration

Registration Request Order

Registration Request Order



since the computing module (CM) is aware of the mobile station’s location during a call via the call origination and termination procedures.



Handoffs 5A mobile station can execute a handoff while it is either in the Idle State or while it is in a call.

An idle handoff occurs while a mobile station has moved from the coverage area of one base station into the coverage area of another base station while it is in the Idle State. If the mobile station detects a Pilot Channel signal from another base station that is sufficiently stronger than that of the current station, the mobile station determines that an idle handoff should occur.

The other four types of handoff occur while the mobile station is in a call.

Figure 5-21What is Ec/Io


What is Ec/Io?

■ Ec/Io• Measures the “strength” of the pilot

• Foretells the readability of theassociated traffic channels

• Guides soft handoff decisions

• Is digitally derived as the ratio ofgood to total energy seen by thesearch correlator at the desired PNoffset

• Never appears higher than Pilot’spercentage of serving cell’stransmitted energy

• Can be degraded by strong RF fromother cells, sectors

• Can be degraded by noise

Ec/Io dB

-25 -15 -10 0

Ec

Io

Energy of desired pilot alone

Total energy received



Figure 5-22What’s in a handset?

Figure 5-23CDMA handoffs

ReceiverRF SectionIF, Detector

TransmitterRF Section

Vocoder

DigitalRake Receiver

Traffic CorrelatorPN xxx Walsh xx



Pilot SearcherPN xxx Walsh 0

ViterbiDecoder

CPUDuplexer

TransmitterDigital Section

Long Code Gen.

Op

en L

oo

p Transmit Gain Adjust

Messages

Messages

Audio

Audio

Bit Packets

Symbols

SymbolsChips

RF

RF

AGC

Bit Packets

Σ

Handoff is the process by which a mobile station maintains communications with the Mobile Services Switching Center (MSC/BSC), when traveling from the coverage area of one base station to that of another

While in theIdle State Idle Handoff

CDMA-to-CDMA Hard Handoff

Softer HandoffDuringa Call

Soft Handoff

CDMA-to-Analog Hard Handoff

While in theIdle State

While in theIdle State Idle Handoff

CDMA-to-CDMA Hard Handoff

Softer HandoffDuringa CallDuringa CallDuringa Call

Soft Handoff

CDMA-to-Analog Hard Handoff



Figure 5-24CDMA soft handoff mechanics

Soft handoff: This is a handoff in which the mobile station starts communications with a new base station without interrupting communications with the old one. Soft handoff can only be used between CDMA channels having identical frequency assignments. Soft handoff provides diversity of Forward and Reverse Traffic Channel paths on the boundaries between base stations.

Soft handoff, in addition to reducing dropped calls, improves their quality. When the mobile station is in soft handoff, the two or three Forward Traffic Channels it receives contain identical modulation symbols with the exception of the power control subchannel. The mobile station provides diversity by combining these different Forward Traffic Channels.

The requirements for a Soft Handoff are the following:

• all links must be on the same CDMA frequency

• all links must use the same traffic frame offset (frame staggering)

• participating BTSs must be connected to the same BSC

■ CDMA soft handoff is driven by the handset• Handset continuously checks available pilots• Handset tells system pilots it currently sees• System assigns sectors (up to 6 max.), tells handset• Handset assigns its fingers accordingly• All messages sent by dim-and-burst, no muting!

■ Each end of the link chooses what works best, on aframe-by-frame basis!• Users are totally unaware of handoff

Handset Rake Receiver

RFPN Walsh

PN Walsh

PN Walsh

SearcherPN W=0

ΣVoice,Data,

Messages

Pilot Ec/Io

BTS

BSCMTX

BTS

Sel.



Figure 5-25Softer handoff

Softer handoff: This is a special case of soft handoff between two or three sectors of the same base station. The DMS-MTX is aware of the softer handoff but does not participate. All the activities are managed by the BTS.

■ Each BTS sector has unique PN offset & pilot■ Handset will ask for whatever pilots it wants■ If multiple sectors of one BTS simultaneously serve a

handset, this is called Softer Handoff■ Handset is unaware, but softer handoff occurs in BTS

in a single channel element■ Handset can even use combination soft-softer handoff

on multiple BTS & sectors

Handset Rake Receiver

RF

PN Walsh

PN Walsh

PN Walsh

SearcherPN W=0

ΣVoice,Data,

Messages

Pilot Ec/Io

BTS

BSCMTX

Sel.



Figure 5-26Overall handoff perspective


Overall Handoff Perspective

■ Soft & Softer Handoffs are the best• but a handset can receive BTS/sectors

simultaneously only on one frequency • all involved BTS/sectors must connect to a single

BSC (the BSC must choose packets each frame)• must be same on all BTS/sectors

■ If above not possible, handoff still can occur but will be “hard” like AMPS/TDMA/GSM

• intersystem handoff: hard• change-of-frequency handoff: hard• CDMA-to-AMPS handoff: hard, no handback

− auxiliary trigger mechanisms available



Figure 5-27CDMA-to-CDMA hard handoff

■ Between cells that could be on the same frequency and have the sameframe alignment, but which are subordinated to different BSCs whichare not interconnected.

• This type of hard handoff would become a soft handoff if the framesreceived at both cells could be delivered quickly to the same BSC forcomparison, by interconnecting both BSCs with a high-speed link(see Inter BSC Soft Handoff / Inter System Soft Handoff)

MTX

PSTN

BSC

MTX

BSC

BA

■ Between cells operating on different frequencies

■ Between cells with traffic channels whose frames are staggered differently

A(ƒ1)

MTX

BSC

PSTN

B(ƒ2)

A(ƒ1)

MTX

BSC

PSTNPSTN

B(ƒ2)

A(∆1)

MTX

BSC

PSTN

B(∆2)

A(∆1)

MTX

BSC

PSTNPSTN

B(∆2)



CDMA-to-CDMA hard handoff take place in any of the following situations:

• The mobile station is transitioning between two cells operating on different CDMA frequencies.

• The mobile station is transitioning between two cells operating on the same CDMA frequency but the traffic channels assigned in both cells have their frames aligned differently (different traffic frames staggering).

• The value of certain parameters defining the quality of the communication goes below configured thresholds and other predefined conditions are met (see Enhanced Hard Handoff triggers)

• The mobile station is transitioning between cells which are connected to different MSCs (MSC = DMS-MTX + BSC), whether in the same or in different CDMA systems (called inter- and intra-system handoffs respectively).

CDMA-to-CDMA hard handoffs take place in a “break-before-make” fashion, and typically take from 0.5 to 1 second to complete.

The mobile station is unable to maintain continuous communications with an MSC. It must stop transmitting, adjust its parameters, and restart the transmission.



Figure 5-28Pilot detection trigger - CELL_PILOT_BEACON sectors

This trigger mechanism utilizes the existing soft handoff algorithm in the mobile station to facilitate hard handoffs.

Certain pilots in the region in which hard handoff is desired are identified as “pilot beacons” in the Pilot Database on the SBS Controller. These pilot beacon could be generated by CDMA cells in an adjacent CDMA system/market, or by a Pilot Beacon Unit. As the mobile station travels into the region in which hard handoff is desired, handoff processing is initiated when the mobile reports to the network that the signal strength of one of these beacon cells has risen above the T_ADD threshold, or exceeds the strength of a pilot in the Active Set by T_COMP.

Notice the difference between “pilot beacon” and “Pilot Beacon Unit”: A pilot beacon is simply a cell/sector identified in the Pilot Database with a cell type of “pilot beacon.” Any pilot used to facilitate hard handoff, regardless of the source of this pilot, can be marked as a pilot beacon in the Pilot Database. A Pilot Beacon Unit is a piece of hardware used to generate a pilot signal.

Finally, observe that the mobile station has no knowledge of the “beacon” cell concept (as far as the mobile station is concerned, the beacon cell is just a

The mobile station has no knowledge of the “beacon sector” concept. As far as the mobile station is concerned,

a beacon sector is just the same asany other standard CDMA sector.

Handoff Trigger

CELL_STANDARD CELL_PILOT_BEACON

BTS BTS The mobile station has no knowledge of the “beacon sector” concept. As far as the mobile station is concerned,


The mobile station has no knowledge of the “beacon sector” concept. As far as the mobile station is concerned,


Handoff Trigger

CELL_STANDARD CELL_PILOT_BEACON

BTS BTS

This trigger utilizes the existing soft handoff algorithm in the mobile station to facilitate the hard handoff.

•Certain pilots in the region where hard handoff is desired are identified as CELL_PILOT_BEACON in the Pilot Database of the SBS Controller

•The cell on the left serves the mobile station on frequency f1

•The cell on the right operates in frequency f2 and has a Pilot Beacon Unit that generates a pilot on frequency f1 (or this pilot is generated by a standard CDMA cell of an adjacent system/market)

•As the mobile station travels into the region in which hard handoff is desired, soft handoff processing is initiated when the mobile station reports to the network that the signal of the beacon cell is received with sufficient strength

•SBS software determines that the reported pilot corresponds to a beacon cell, and hard handoff processing commences



standard CDMA cell), and that any CDMA cell or any cell equipped with a Pilot Beacon Unit can be designated as a pilot beacon in the Pilot Database.

Figure 5-29Hard handoff using Beacon Pilot sectors

ƒ1 ƒ1 ƒ2 ƒ2ƒ1

ƒ2 ƒ2

Beacon Sector on ƒ1

ƒ1 ƒ2

Beacon Sector on ƒ2

ƒ1

BEACON

BEACON



Figure 5-30Boundary sector trigger – border cells

This is a two-stage trigger mechanism. The first-stage trigger occurs when the mobile indicates that all the strength of all the non-boundary pilots had fallen below T_DROP (and remained there for at least T_TDROP seconds). In other words, this trigger occurs when the mobile station’s Active Set consists only of pilots from sectors marked as boundary sectors. This type of Active Set is known as a Boundary Active Set.

When the mobile station’s current Active Set is a Boundary Active Set, the second-stage trigger is enabled. Since soft handoff processing is independent of hard handoff processing, the mobile station’s Active Set could change after the second-stage trigger has been enabled. If the mobile station’s Active Set becomes a non-Boundary Active Set prior to the second-stage trigger occurring, the second-stage trigger is disabled.

Unlike the Pilot Detection trigger, with the Boundary Sector trigger, target selection does not begin until this second-stage trigger occurs.

When the second-stage trigger is enabled, the Round-Trip Delay (RTD) of communication with the mobile station is monitored by the SBS. The RTD measurement is used to estimate roughly the distance between the mobile station and the base station. Setting the RTD threshold to an adequate value

■ This is a two-stage trigger which indirectly utilizes the existing soft handoff algorithm in the mobile station to facilitate the hard handoff

• certain pilots in the region where hard handoff is desired are identified as “CELL_BORDER” in the Pilot Database of the SBS Controller

• as the mobile station travels from left to right, it enters into handoff with both sectors and eventually ceases communication with the sector on the left (“CELL_STANDARD”)

• when the active set contains only sectors datafilled as “CELL_BORDER,” the first-stage trigger is met and the second-stage trigger is enabled

• the SBS starts monitoring the Round Trip Delay (RTD) of the signals between the mobile station and the base station from which it derives its time reference

• when the RTD exceeds a certain threshold, the second-stage trigger is met and handoff processing continues with the target selection activity

The mobile station has no knowledge of the “border sector” concept. Asfar as the mobile station is concerned, a border sector is just the same asany other standard CDMA sector.

First-stage trigger

Second-stagetrigger

CELL_STANDARD CELL_BORDER

RTDBTS The mobile station has no knowledge

of the “border sector” concept. Asfar as the mobile station is concerned, a border sector is just the same asany other standard CDMA sector.

First-stage trigger

Second-stagetrigger

CELL_STANDARD CELL_BORDER

RTDBTS



maximizes boundary sector coverage by preventing hard handoff until the mobile station is a certain distance from the base station. Only when the RTD exceeds a certain threshold (datafilled in the Pilot Database), does handoff processing continues with the target selection phase.

The RTD threshold for each boundary sector must be determined through drive testing. It should be set such that the hard handoff can be completed while the mobile is still within the coverage of the boundary sector. Therefore, the RTD is different for different propagation environments and traffic speeds.

Summarizing: the boundary sector hard handoff trigger is the state in which the mobile station is in communication only with cells whose pilots are designated as boundary pilots and the RTD of communications with the mobile station exceeds a certain threshold.

Notice that for the Boundary Sector Trigger mechanism to work, both the boundary sectors and the sector/cells from where the mobile is exiting must be served by the same BSC.

Notice also that this handoff trigger could fail to operate in certain environments. For instance, if a boundary sector has an Active Set which includes a pilot from a distant cell (due to some propagation anomaly), the boundary sector hard handoff trigger does not occur; the Active Set in this case is not a Boundary Active Set, although the mobile station is physically in a boundary sector. The mobile station will “drag” the boundary sector and runs the risk of dropping the call before a Boundary Active Set exists.



Figure 5-31Hard handoff using border sectors

ƒ1 ƒ1 ƒ2 ƒ2

ƒ1 ƒ1 ƒ2 ƒ2

Border Sector for ƒ1

Border Sector for ƒ2



Figure 5-32CDMA-to-analog handoff

CDMA-to-analog hard handoff: This is a handoff in which the mobile station is directed from the Forward Traffic Channel to an analog voice channel of the 800 MHz analog system.

Intra-system CDMA-to-analog handoff In many cases, the CDMA network is installed over an existing AMPS cellular network. In such cases the CDMA cells/sectors are often overlaid on top of the AMPS cells/sectors. It is possible that under certain situations, the RF link quality of a call on the CDMA network may be unacceptable. To avoid a potential call drop, a handoff to the AMPS network is initiated. The AMPS network may be able to support the call, thus saving a call drop. This feature facilitates a handoff trigger, from CDMA to AMPS, if the RF link quality of the cell on CDMA falls below a certain threshold.

In order for the handoff from CDMA 1900/800 MHz to 800 MHz AMPS to occur, the CDMA system should be overlayed on an existing AMPS system one-to-one. When one CDMA cell/sector overlays exactly one AMPS cell/sector then it is referred to as one-to-one overlay, and this the target selection unambiguous.

■ The mobile station is directed from a forward traffic channel to ananalog voice channel

■ Radio link continuity is not maintained■ Two types of handoff:

• Inter-system - occurs while the mobile station is traveling intoanother system that has no CDMA service

− Messaging will tell the mobile station to select AMPS− Currently, the mobile station cannot handoff back from

AMPS to CDMA (until the end of the call, when the mobilestation reacquires the system) because the necessarysignaling messages not supported)

• Intra-system - occurs while the mobile station is traveling withinthe system

− Load balancing− Improve voice quality− No CDMA service



Enhanced hard handoff triggersThe Enhanced Hard Handoff is designed to be triggered under the following conditions:

• AMPS provides in-building coverage where CDMA may not.

• Holes in the core CDMA coverage

• As a result of getting off a highway and onto a side road which no longer has CDMA coverage

• Small Core Area with only a few CDMA cell sites

The CDMA cell is overlaying the AMPS cell. The CDMA sector in the region in which Hard Handoff is desired are identified as “Link Quality” sector in the Pilot Database.

Initially the Mobile Station (MS) is communicating through a CDMA link, as the MS moves away from the CDMA cell site the RF link becomes too weak to support the communications at an acceptable level of quality.

Threshold for certain quality-related parameters are set depending on the minimum link quality desired to maintain an acceptable voice quality. When this threshold is reached the call could potentially drop. To avoid a call drop, a “link quality hard handoff” (also called Enhanced Hard Handoff) is triggered. As a result a new communication path is established through an AMPS link, and the mobile is re-directed to that link using the existing hard handoff functionality. From this moment, and until the end of that call, the MS communicates with the AMPS cell site.

If the CDMA link quality improves before the mobile is asked to switch over to the AMPS sector, the handoff is abandoned. This may be referred to as a “handoff saved”. The target AMPS system would time-out and release its resources that were set to receive the call if the handoff had occurred.





1. What is the purpose of power control?

2. What is the ideal situation at the base station regarding power level of the received mobile station signals?

3. Define CDMA registration and explain its purpose.

4. Name the forms of registration, indicating whether they are autonomous or not.

5. Identify the cases of CDMA handoff that can occur when the mobile station is in the Traffic Channel State.

6. Identify the message sent by the mobile station to report the strength of the pilots it measures

Family Product Manual Contacts Copyright Confidentiality Legal statements DocInfo

CDMA

Technology OverviewStudent Guide

To order Wireless Solutions documentation, call 1-800-NTI-CARE (1-800-684-2273)

To report a problem in Wireless Solutions documentation, call 1-800-NTI-CARE (1-800-684-2273)or send e-mail from the Wireless Solutions Training and Documentation World Wide Web site at http://www1.nortelnetworks.com/wireless/DocuTrain/

Copyright 1996–2000 Nortel Networks Corporation, All Rights Reserved

NORTEL NETWORKS CONFIDENTIAL

The information contained herein is the property of Nortel Networks and is strictly confidential. Except as expressly authorized in writing by Nortel Networks, the holder shall keep all information contained herein confidential, shall disclose it only to its employees with a need to know, and shall protect it, in whole or in part, from disclosure and dissemination to third parties with the same degree of care it uses to protect its own confidential information, but with no less than reasonable care. Except as expressly authorized in writing by Nortel Networks, the holder is granted no rights to use the information contained herein.

Information is subject to change without notice. Nortel Networks reserves the right to make changes in design or components as progress in engineering and manufacturing may warrant.

* Nortel Networks, the Nortel Networks logo, the Globemark HOW the WORLD SHARES IDEAS, and Unified Networks are trademarks of Nortel Networks. Trademarks are acknowledged with an asterisk (*) at their first appearance in the document.Course number: Course 809AProduct release: NBSS9.0Document version: Standard 04.02Date: October 2000Printed in the United States of America

Date post:	12-Oct-2014
Category:	Documents
Upload:	antar-almobarezi
View:	79 times
Download:	1 times

CD Ma Student Nortel

Documents