Low-Power Correlation for IEEE 802.16 OFDM Synchronization

on FPGA

By S.M.Brundha (10G21A0455)D.Navajyothi (10G21A0426)I.Akkil (10G21A0438)MC.Chaithanya(10G21A0419)

ASCET

• Aim • Objective • Motive • Existing system• Proposed system • Implementation flow• Operational flow• Simulation and synthesis report• Applications • Conclusion

CONTENTS

To implement a low-power correlator for IEEE 802.16 OFDM synchronization on

FPGA

AIM

The main objective of implementing this

correlator is that it consumes less power than the

correlator of the existed technology

OBJECTIVE

MOTIVE

• Here IEEE 802.16 refers to the wireless Metropolitan area network(WMAN)or wireless microwave access(WIMAX) and the multiplexing technology we use is OFDM

• It is necessary to have timing synchronization for the transmitted signal so we use cross correlation technique for the received samples and filter coefficients.

• Initially it is implemented with the multipliers later multiplier less also in the multiplier less we replace the adders with the low power one

WIMAX

• where as WLAN (IEEE 802.11) restricted to smaller area(100m) • It has lower data rate(54Mbps)• WLAN referred to as WIFI provides low quality of service

• IEEE 802.16d used here refers to the fixed microwave access used for broadband internet access

• Broadband refers to larger channel band width provided for

larger data rate (100Mbps)

• WIMAX technology can be extended for larger area (10Km)

•By using a single base station in an area we can serve many users in different areas.

WLAN Service

WMAN service

OFDM TECHNOLOGY• It is an extension for FDM • we can serve more number of customers using a single

channel where as in FDM a single channel will render service to one customer.

• Efficient channel usage in OFDM with no guard band by making signals to be orthogonal

Cylic prefixing

CORRELATION

• Here we use cross correlation this happens between the received samples and between the coefficients

• This is applied to find whether there is timing synchronization for the received signal i.e whether the signal reached receiver on time

• Generally correlation is done to attain the original signal from the noise effected one’s based with the preamble symbols

Multiplier based cross correlation

• In the previous cases multiplier based correlation is used and It is complex to implement .

• Correlation is similar to FIR filter implementation we go in similar to DSP implementation

As in general we have for FIR filters

output=ax(n)+bx(n-1)+cx(n-2)+dx(n-3)+…..

Similarly for correlators after Z transformOutput = Pr [63] Ri + z−1( Pr [62] Ri + z−1( Pr [61] Ri + · · ·)

Transpose direct form of the correlator

Eg:Rin*pr[63]=[1010]*[0]=0000Rin*pr[62]=[1010]*[1]=1010 and so on

•In the multiplier less based technology we are going to use the shift and add technique initially the input along with the preamble samples are being shift and added •The out put will be of 2 types with carry or without carry which is provided to the multiplexar based on the preamble symbol which acts as a selection line the required output is attained and it is is given to the adder • • Atlast all the output of the adder are added and then we attain the required correlation for timing synchronisation

Eg:Rin=1110 If preamble bit is 0 then

Multiplier less Implementation

•In the multiplier less based technology we are going to use the shift and add technique initially the input along with the preamble samples are being shift and added •The out put will be of 2 types with carry or without carry which is provided to the multiplexar based on the preamble symbol which acts as a selection line the required output is attained and it is is given to the adder • Atlast all the output of the adder are added and then we attain the required correlation for timing synchronisation•

• In the existing technology of WIMAX we have used the carry save adder for the correlation operation for timing synchronization which may consume more area as we have a separate block for storing the carry which consumes more power and area.

EXISTING TECHNOLOGY

• In the carry save adder addition initially all the bits are added we attain a carry and sum at each addition from right to left the carry attained is not added with the next addition instead stored and at last for all the sum bits we add the carry bits

CARRY SAVE ADDER

There is one CSA unit per bit• Every CSA has a c output and an s output• Doing this way, addition can be implemented in log m sums• It is necessary to implement one final Adder All these consumes extra area for the register used for carry saving and consumes more power for the operation

• Hence we are going to replace the existing technology correlator’s carry save adder with the ripple carry adder hence thereby the power consumed decreases and the area occupied also decreases

PROPOSED TECHNOLOGY

Ripple carry adder

In the previous case the carry save adder’s drawback is overcame by using the ripple carry added it doesn’t require any extra register for the storage of the carry bits

The carry bit is directly added with the sum bit of the next addition

The layout of ripple carry adder is simple, which allows for fast design time

however, the ripple carry adder is relatively slow, since each full adder must wait for the carry bit to be calculated from the previous full adder. The gate delay can easily be calculated by inspection of the full adder circuit.  

IMPLEMENTATION FLOW

OPERATIONAL FLOW

1 1

1

Softwares Utilised

• we require the usage of 2 softwares

Modelsim and Xilinx ISE Code is written on xilinx ISE 10.1 editor

window The simulation is being done by using

Modelsim 6.4b

MODELSIM SIMULATION

SYNTHESIS REPORT

RESULTSpartan3e-xc3s250

LUT SLICES DELAY Power

CSA 4404 2353 8.545ns 6.128 W

RCA 3425 1869 6.800ns 5.962 W

HARDWARE UsedThe code is being dumped in to the FPGA SPARTAN 6

FPGA are reprogrammable they use HDL for programming These are flexible we can use it for digital signal processing, as a microcontroller ,micro processor by programming the device to achieve the required operation

CLB Contains look up tables refers to group of logic gates forming an boolean equation, flipflops, multiplexars .

The CLB’S are connected to any of the wire segment similarly I/O blocks I/O Blocks are used for read and write operation

• Wimax used for broadband internet access for over larger distance and with high speed

• consumers, businesses, and government agencies alike will discover a whole new broadband paradigm, which will significantly enhance their ability to conduct their business

APPLICATIONS

CONCLUSION

Over the multiplier based implementation multiplier less consumes less power also replacing carry save adders with the ripple carry adders we can attain less power usage .

An additional benefit of multiplierless correlation is that it can be used on any FPGA architecture.In this RCA implementaion we can achieve less delay overall than with CSA

REFERENCES• IEEE Xplore Abstract - Low-Power Correlation for IEEE 802.16 OFDM Synchronization on

FPGA - http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=6295680#

• Intel® My WiFi Technology –http://www.youtube.com/watch?v=5DHBFrPBNc8• WiFi Vs WiMax -http://www.youtube.com/watch?v=Y7ydkmFaP7E

• EEVblog #496 - What Is An FPGA? –http://www.youtube.com/watch?v=gUsHwi4M4xE

• Finite impulse response --http://en.wikipedia.org/wiki/Finite_impulse_response

• Understanding and Using Correlation Analysis TradeStation Labs-https://www.tradestation.com/education/labs/analysis-concepts/understanding-and-using-correlation-analysis

• BBC - GCSE Bitesize: Communicating with waves: radio waves and microwaves-http://www.bbc.co.uk/schools/gcsebitesize/science/aqa/waves/soundandlightrev3.shtml

• SSCE-Shift-Mult.pdf-http://users.utcluj.ro/~baruch/book_ssce/SSCE-Shift-Mult.pdf

• Inter Carrier Interference (ICI) in OFDM-http://www.dsplog.com/2009/08/08/effect-of-ici-in-ofdm/

• What is difference between OFDM and FDM-http://www.edaboard.com/thread169470.html

• Orthogonal frequency-division multiplexing - Wikipedia, the free encyclopedia-http://en.wikipedia.org/wiki/Orthogonal_frequency-division_multiplexing#Orthogonality

• Cross-correlation – ALGLIB-http://www.alglib.net/fasttransforms/crosscorrelation.php

• Cyclic prefix - Wikipedia, the free encyclopedia

• Carry save addition-people.clarkson.edu/~linh/algorithms/carry%20save%20addition.pdf

• Ripple carry adders-www.engr.uky.edu/~jel/misc/d481/info/adders/adders.html