9500 MPR Release 5.1 Alcatel-Lucent 9500 Microwave Packet Radio (MPR) is a solution for smooth transformation of backhaul networks from TDM/ATM to Ethernet. The 9500 MPR solution efficiently transports whatever multimedia traffic since it handles packets natively (packet mode) while still supporting legacy TDM traffic (hybrid mode), with the same Hardware. It also provides the Quality of Service (QoS) needed to satisfy end-users. This solution not only improves packet aggregation, but also increases the bandwidth and optimizes the Ethernet connectivity.
Transcript
9500 MPR Release 5.1
Alcatel-Lucent 9500 Microwave Packet Radio (MPR) is a solution for smooth transformation of backhaul networks from TDM/ATM to Ethernet. The 9500 MPR solution efficiently transports whatever multimedia traffic since it handles packets natively (packet mode) while still supporting legacy TDM traffic (hybrid mode), with the same Hardware. It also provides the Quality of Service (QoS) needed to satisfy end-users. This solution not only improves packet aggregation, but also increases the bandwidth and optimizes the Ethernet connectivity.
An embedded 10 Gbit/sec L2 switch is present on the card.
There are 4 Electrical 10/100/1000 base-T electrical ports and 4 optical SFP (LX and SX).
Supported features:
• IEEE 802.1D
• User Selectable QoS : none, DiffServ or 802.1p bits
• VLAN management 802.1Q
• Q-in-Q IEEE 802.1Q
• Port segregation
• Flow control 802.3x
• Auto-negotiation enable/disable
• Support of jumbo frames (9728 bytes) on FE/GE interfaces
• Per port policer
• Per flow policer
• Broadcast/Multicast storm control
• MAC address control list
• VLAN swap
EAS card can be used optionally as interface card to interconnect up to 4 MPTs; supporting up to 24
MPT with a single MSS8.
Additionally, EAS card supports Multichannel LAG L1 feature. Multichannel feature provides a
solution where more traffic capacity is needed than can be transported over one physical link. N
25
radio links are aggregated to provide one logical link with a capacity that is the sum of the individual
links. This feature is particularly useful for wireless transmission systems where multiple radio links
must be used in parallel to achieve very high capacities of 1Gbit/s and above. This provides optimum
payload balance, regardless of the throughput demands of individual user connections
Redundancy is also a feature of multichannel aggregation. If a link is lost, its traffic is directed onto
the remaining link(s) within the group.
If the Ethernet bandwidth on the remaining link(s) is over-subscribed, traffic will be dropped, though
with appropriate QoS settings only low priority data will be affected - all high priority data will
continue to get through.
Multichannel feature can be applied in principle to any kind of traffic: Ethernet, TDM, ATM and SDH.
Multiline feature is supported by EAS 8 Gbit/card, with MPT-HC connected to optical ports.
LAG groups can be IntraEAS (all MPTs on same EAS card) or CrossEAS (MPT on EAS on the same Row);
here below some example of supported configuration. Maximum number of MPTs in a LAG group is
4.
Core NE A
EAS
Core NE B
EAS2EAS1
4 RFChannel
s
4 RFChannel
s
rLAG1 rLAG1 rLAG2
EAS2EAS1StackingrLAG1
rLAG2
Core NE A
EAS2
rLAG1 rLAG1
rLAG2
Core NE B
EAS1
rLAG1
4 RFChannel
s
4 RFChannel
s
Stacking
26
Electrical and Optical EAS ports not belonging to a LAG can be used as User Ports or Radio Interfaces
(SFP ports only) both in 1+0.
Aggregated Radio Links should have same modem profiles.
Adaptive Modulation, ring protection has been introduced in conjunction with multichannel.
5.4 2E1 SFP In order to target applications where a few number of E1s are needed, a miniature E1 over GE
converter is available. 2E1 SFP is SFP device that provides two G. 703 E1 interfaces, supporting the
same functionalities of 32E1 PDH card. In addition, this device is able to generate a “dummy framed”
E1 in order to provide synchronization to an external equipment (like a BTS).
This device can be used instead of 32E1 PDH card when the requested E1 connectivity is limited,
saving in this way one slot in MSS4/MSS8 that can be used by other cards.
2E1 SFP
2xE1 SFP can be plugged in one of the two SFP ports of Core card, providing two G. 703 E1 interfaces
(up to 4xE1 in case Core Card hosts 2 SFP). EPS protection is available in case Core Card is protected:
the secondary SFP is hosted by the stand-by Core, and a Y cable is provided to connect the 2 SFP.
27
5.6 SDH Access Card
9500MPR SDH Access card is the board that enables 9500 MPR to be connected to a SDH network.
The same board can be used in two different working modes, addressing two different network
scenarios:
• STM-1 mux/demux
• STM-1 transparent transport over the radio
SDH Access Board
28
5.6.1 STM-1 mux/demux application
The STM-1 mux/demux behaves as a terminal multiplexer; it terminates or originates the
SDH frame. It multiplexes up to 63xE1 into a STM-1 electrical/ optical line connection.
Standard VC4 mapping of lower-order E1 traffic streams to/from STM-1 is applied, that
means that a VC4 directly maps up to 63xVC12 into an STM-1 signal (in turn each VC12
contains 1xE1)
Typical application is a direct connection to SDH add-drop multiplexers (ADMs)
5.6.2 STM-1 transparent transport application
In this application the board has the aim to manage the specificities of the related external
interface and to implement the adaptation function between the external interface and the
boundary internal interface. Up to 2xSTM-1/OC-3 are transparently transported through a
single radio link.
The card supports 1xSTM-1 in channelized mode or up to 2xSTM-1 interfaces in transparent
transport mode (2 optical interfaces or 1 electrical interface)
The Front Panel Interfaces include:
• 2x SFP (optical LC connector or electrical 1.0/2.3 connector)
• One Led indicator for status
In case of EPS line protection two boards are plugged inside the sub rack. Optional splitter Y-cables
are provided for both Tx and Rx SDH signals.
29
5.7 EoSDH SFP
Ethernet over SDH (EoSDH) SFP is miniature Gigabit Ethernet over STM-1/OC3 converter that bridges
between GE networks and SDH networks providing simple and efficient Gigabit Ethernet connectivity
over SDH.
The device offers a migration path for connecting future-ready IP devices to existing SDH/SONET
networks
EoSDH SFP
EoS SFP supports the following basic features:
Delivers Gigabit Ethernet traffic over a single STM-1/OC-3 link
Supports standard GFP encapsulation according to G.7041/Y.1303: Gigabit Ethernet frames are
mapped into VC-4 or STSc-3
Physical interface is 1xSTM-1 optical in a SFP cage with LC connector.
EoSDH SFP can be plugged in one of the two SFP ports of Core card (up to 2xSTM-1 in case Core Card
hosts 2 SFP). EPS protection is available in case Core Card is protected: the secondary SFP is hosted
by the stand-by Core, and an optical splitter is provided to connect the 2 SFP.
30
5.8 E3 SFP
E3 SFP is a TDM Pseudo wire access gateway extending TDM-based services over packet-switched
networks.
E3 SFP
The device converts the data stream from its user E3 interface into packets for transmission over
9500 MPR network; the addressing scheme is MEF8. These packets are transmitted via the SFP port
of the Core Board; a remote E3 SFP converts the packets back to TDM traffic.
Physical interface is 1xE3 electrical in a SFP cage with 1.0x2.3 connector.
E3 SFP can be plugged in one of the two SFP ports of Core card (up to 2xE3 in case Core Card hosts 2
SFP.
EPS protection is available in case Core Card is protected: the secondary SFP is hosted by the stand-
by Core, and a Y cable is provided to connect the 2 SFP.
31
5.9 MPT Access Card
The MPT Access Card is dedicated to connect the MPT to MSS,.
Up to two MPT can be connected to the MPT Access Card
Main physical characteristics:
• 2 x 10/100/1000 Base – T Port for electrical data to/from MPT. These ports can also
power the MPT through the same CAT5 cable.
• 2 x SFP Optical GETH for optical data connectivity to/from MPT
• Double 50Ω QMA Connectors as an option for MPT Power feeding in case of optical
connectivity
Main Functions:
o Provide traffic interface between Core switch and MPT
o Provide the power supply interface to the MPT
o Lightning and surge protection for both electrical GETH and power interfaces that are
connected to MPT
o MPT 1+1 protection management
o Clock distribution function
o Radio Link Quality notification through MPR Protection Protocol frames
o Communication with Core controller for provisioning and status report.
MPT Access Card
32
5.10 Power injector plug-in
This card can be used for several applications:
• When MPT is connected to CORE, power injector is needed to provide power to the MPT
at optimized price • When MPT is used in stand alone (MPR-e) and connected to 7705SAR, Power injector
plug-in can be used inside 7705 chassis to power MPT
A box version is also available for all other applications of MPR-e.
Main physical characteristics:
• 2 DC connectors in the front (box), or power from the backpanel. • 2 RJ45 for the data in • 2 RJ 45 for the data + DC out • 2 LEDs indicating the presence of DC voltage on each Ethernet output
Power injector plug-in
33
5.11 AUX board
Service channels accesses and housekeeping alarm are supported by auxiliary peripheral.
Auxiliary cards support two main functions:
• Auxiliary data channels management (2 x 64 Kbit/s service channels)
• External I/O management
AUX Board
Auxiliary board front panel is equipped with four connectors:
• EOW connector
• Service channel interface #1 (RS422 V11 DCE 64 kbit/s)
• Service channel interface #2 (RS422 V11 DCE 64 kbit/s)
• Housekeeping interface (6 inputs + 7 outputs. The polarity of each alarm is user configurable
and a user defined label could be added per each alarm)
Only one auxiliary card per NE can be equipped, and in a fixed position: it can be lodged in slot 8
(bottom right) of MSS-8 or in slot 4 (bottom right) of MSS-4.
Typical applications for AUX board are :
• transport over MPR of the ingress service channels that could be delivered for example by
9400 LUX 40/50, LUX12, 9400AWY 2.0/2.1, 9500 MXC
• transport over MPR of the ingress service channels that could be delivered by end user. Note
in case of 64 Kbit/sec the end user must be always configured as DTE.
34
• transport over MPR of the TMN signal coming from:
o LUX 12, V11 9.6 Kbit/s RQ2 protocol
o LUX 40/50, V11 9.6 Kbit/s SNMP protocol
Please note that in the last case MPR is taking care of pure transport; no termination of TMN channel
is done inside MPR using aux card, while recommended TMN chain is done using Ethernet TMN
interface for 9400AWY and 9500 MXC.
35
5.12 Fan Board A FAN card is required inside the MSS-4/8 shelf. MSS-4 can be optionally equipped without fan card,
supporting temperature up to +45°C 1. The FAN holds three long-life axial fans, which are controlled
and performance-monitored by the controller.
Fan Board
To have high reliability 3 fans are used with separate alarms in order to understand the urgency (two
or three fans failed) or the not urgency condition (one fan failed).
The Unit is inserted from front side to avoid payload interruptions in case of fan maintenance. The
FAN is hot swappable and in-service replacement doesn't affect traffic.
An optional Fan unit, called Fan Alarm Card, is available on MSS-8, hosting a housekeeping connector
for Equipment Alarms (Summary, Major and Minor) and 4 housekeeping inputs and 8 high reliability
fans. The board is mandatory when 24V DC converter is equipped.
36
5.13 +24V integrated DC/DC converter
An optional +24V DC/DC converter is available for MSS-8 shelf
One or two converters are able to slide on the MSS chassis, side by side, in a single card slot.
Unprotected converter kit will be used in configurations where single, non –redundant “A” battery
feed is used. Protected converter kit will be used when dual, redundant, “A” and “B” battery feeds
are used. In either configurations, the +24VDC to -48VDC converter kits use a single vacant slot of the
MSS chassis.
There is no interconnection between the converter(s) and the MSS backplane. Both the +24 VDC
input and -48 VDC output are available via 2 position connectors on the front of the unit.
The converter(s) will receive its input(s) from +24 VDC primary power feed(s) and the -48 VDC
output(s) will be connected to the MSS -48 VDC inputs located on the right side of the MSS chassis
via a short external power cable, providing -48 VDC to the MSS, in the same way the shelf is powered
when -48 VDC primary is used as oppose to +24 VDC.
+24V DC/DC converter can power any module in the shelf (and of course related ODU connected to
the module) up to a total power consumption of 348 watts.
When + 24V DC/DC converter is used, the Fan Alarm board must be equipped in the rack.
37
6 MEF-8
6.1 MEF-8
As described in MetroEthernet Forum, MEF-8 is a standard for “implementing interoperable CES
equipment that reliably transport TDM circuits across Metro Ethernet Networks while meeting the
required performance of circuit emulated TDM services as defined in ITU-T and ANSI TDM
standards”. The Circuit Emulation Service (CES) emulates a circuit network, by packetizing,
encapsulating and tunneling the TDM traffic over Ethernet.
MEF-8 Service Definitions
Alcatel-Lucent 9500 MPR implements a proprietary technique that reduces to a few percentages the
overhead improving the use on bandwidth on air when MEF-8 emulated circuits are transported. The
improvement depends on the MEF-8 payload size and frame format and in case of TDM2TDM results
in having quite the same efficiency than a traditional TDM radio.
6.1.1 BER performances
When MEF-8 Ethernet frames are transmitted through a noisy medium (e.g. the Radio Physical
Layer), bit errors may occur. If an Ethernet frame is affected by one error, this is detected and the
entire frame is dropped. This affects the TDM with a worse BER that if compared with a traditional
TDM over TDM transmission process, it is higher, multiplied by a factor that is the frame length.
In order to avoid such BER degradation a technique is implemented such as for any reasonable BER
on the Radio Channel, the TDM transported by MEF-8 CESoETH is affected by the same BER without
any multiplication effect.
38
6.1.2 Packet Delay Variation control
A technique is implemented in order to control Packet Delay Variation (PDV) affecting MEF-8
Ethernet frames. With this technique the waiting time that affects MEF-8 Ethernet frames are not
depending on the length of the Ethernet frame.
This gives benefit in term of packet delay variation minimization, so that any kind of services (VoIP,
TDM, ATM, Ethernet) is experiencing a small cost value of PDV, independently and regardless of the
traffic load.
7 Adaptive Modulation
To be able to fulfill the required quality of service (QoS) parameter of the specific applications,
together with the goal of efficient usage of the available frequency spectrum under temporal
variable channel conditions, the signal transmission parameter should be adapted to the near-
instantaneous channel conditions.
The receiver measures/estimates the communication channel conditions and sends a report to the
transmitter station. The signal transmission parameters are determined for the next transmission
according to channel quality estimation. The transmitter and the receiver must regularly synchronize
the applied communication mode.
An appropriate prediction method is needed for channel parameter estimation, because channel
quality estimation error limits the performance of the adaptive system. The most reliable approach is
based on the Signal-to-Interference-plus-Noise-Ratio (SINR), measured obtained using the Mean
Square Error (MSE).
The radio with ACM is "error-less", in other words is able to guarantee the same performances
either in case of Constant Bit Rate (CBR) payload or in case of "First Priority" payload. The error-less
concept means that a certain portion of the traffic, i.e. SDH, PDH or other-like CBR or NCBR defined
by the customer/operator as "first priority", shall be treated as the traditional traffic in SDH or PDH
system, guarantying a certain level of availability.
The remaining portion of traffic is carried with less availability, according to the link propagation
performances, guarantying the "best effort" or other objectives.
39
9500 MPR allows to fully exploit the air bandwidth in its entirety by changing modulation scheme
according to the propagation availability, associating to the different services quality the available
transport capacity.
7.1 Performances of Adaptive Modulation: • for Flat Fading, 9500 MPR supports notch speed up to 100 dB/sec without errors on priority
traffic.
• in case of Selective Fading 9500 MPR is able to provide a 40 dB notch event, thus supporting
100 MHz/sec speed without errors.
40
8 Synchronization
The Alcatel-Lucent 9500 MPR product family supports a full range of local and end-to-end network-
synchronization solutions for a wide variety of applications.
At the ingress of the microwave backhauling the network clock can be locked to anyone of the
following sources:
• Synch-Eth
• Any plesiochronous E1/T1 data link chosen from any input interface
• Dedicated Sync-In port available on MPR core module for a waveform frequency signal at 2,
5, or 10 MHz
• Built-in free run oscillator.
• STM1 clock chosen from SDH input interface
At the egress of the backhauling network synchronization is made available through anyone of the
following:
• Synch-Ethernet according to G.8261/8262
• Any plesiochronous E1/T1 data link chosen from any output interface
• Dedicated Sync-In port available on MPR core module for a waveform frequency signal at 2,
5, or 10 MHz.
• STM1 clock chosen from SDH output interface
It is important to notice that ingress and egress methods can be freely mixed, depending on the
specific needs of the operator. So, as an example, the network clock can be locked to an ingress E1
and delivered through a Synch-Eth or BITS interface at the egress of the microwave backhauling.
On the radio channel, a 9500 MPR transfers the reference clock to an adjacent MPR device through
the radio carrier frequency at physical layer. This method offers two main advantages:
• No bandwidth is consumed for the synchronization distribution
• Total immunity to the network load.
End-to-end scenarios where time-of-day/phase alignment are requested are fully supported, as 1588
PTP v2 is delivered transparently by MPR across the microwave backhauling network.
41
PRC
Cell site
Aggregation
network
Cell site
Cell site
Possible synchronization sources: • E1/T1 available for data traffic
• 2.048 MHz, 5 or 10 MHz input
Possible synchronization options: • E1/T1 • 2.048, 5 or 10 MHz output
RNC
Synchronization distribution path
Point of availability of the synchronization
1588 transparent transport
MPR deployment in mobile backhauling
Both for Hybrid and Packet working modes, the Clock can be received at hand-off or delivered at the
cell site. Synch-Eth, E1, PDH, SDH and BITS clock modes are available.
9500 MPR has an embedded reference clock which is distributed to each board of the network
element. Such clock is generated in the Clock Reference Unit (CRU) of the core card (controller).
Clock source selection and distribution
PDH cardPDH card
ASAP cardASAP card
Radio cardRadio card
Core cardCore card
E1/T1
CRUCRUClock
selector
Clock
selector
G813
quality
ATM/IMAE1/T1
Symbol rate
Synch- EthSynch-Out
PDH cardPDH card
ASAP cardASAP card
Radio cardRadio card
Core cardCore card
E1/T1
ATM/IMAE1/T1
Symbol rate
Synch- Eth
Synch-Out
Stratum 3oscillator
Distributed
reference
clock
SDH/Sonetcard
SDH/Sonetcard
STM-1/OC-3
SDH/Sonetcard
SDH/Sonetcard
STM-1/OC-3
42
The availability of the Clock in the Network represents the most common scenario, characterized by
a time source available at the ingress of the microwave backhauling network, derived from the
primary reference clock.
Network Clock Available
Synchronization (frequency) is delivered to the cell site using any of the options available on MPR,
depending on the operator’s need. Worth repeating ingress and egress methods can be mixed (i.e.
Synch-Eth at the ingress, E1/T1 at the egress) via a simple configuration.
PRC
Service nodewith master clock
Microwave tail
Microwavehub
Microwave hand-off
Cell site
Aggregation network
Sync-Eth
T1/E1
BITS
1588
Aggregation
network
L1 synchL1 synchSync-Eth
T1/E1
BITS
Network
clock –
frequency
Network
clock –phase
Service
clock
Sync-Eth
SDH
DCR
43
9 Ethernet Features
9.1 MAC Switching – embedded Level 2 Ethernet The switch is capable to evaluate the destination address of each frame received and to transmit the
individual frames to the correct egress port according to information contained in a database
"address resolution table" and associated to destination address. If the switch does not know on
which port to forward the frame (destination address is not present in "address resolution table"), it
sends the packet on all ports (flooding). The switch performs half transparent bridge functionality
that is to filter the frames which destination is on the segment (port) where it was received.
9.2 Level-2 Addressing The address management function is performed in the switch through the address table (Level-2
Table) that can manage up to 16384 entries in MSS-4/8, 8192 entries in MSS-1c. This means that the
maximum number of MAC addresses supported is 16384 for MSS-4/8 and 8192 with MSS-1c.
New entries are automatically learned when packet is received on port.
These entries can be created or updated by the Equipment.
The aging process periodically removes dynamically learned addresses from the "address resolution
table".
Learning is based on Source MAC Address and VLAN ID.
It is possible to combine this function with the static configuration of the registration entries. For any
valid incoming packet, the Source MAC Address is associated to the VLAN ID (directly from the packet
or through VLAN Tables) and used to search the proper tables.
If a match is not found, the new address is learned and associated with the ingress port of the
packet. If a match is found, no further action is taken for learning.
The Destination MAC Address along with the VLAN ID is used as a search key for the packet’s output
port.
If a match is found then the packet is switched out on the matched port, otherwise, if the match is
not found, then a Destination Lookup Failure (DLF) occurs and the packet is switched out on all ports
that are members of the VLAN, except that one which has received the packet in ingress.
44
9.3 Flooding If the switch does not know on which port to forward the frame (destination address is not present in
"address resolution table"), it sends the packet on all ports (flooding). By default the flooding is
enabled on all ports and doesn’t require any CT/NMS setting. Nevertheless using the cross
connections capability is possible to restrict the flooding only on some ports.
9.4 Half bridge functionality The switch performs half transparent bridge functionality (address learning to filter the frames which
destination is on the segment where it was generated).
9.5 Summary of Ethernet Features Supported
9.5.1 IEEE 802.3x Flow control
In case of incoming Ethernet traffic leading to exhaustion of buffers on input queues, PAUSE frames
are transmitted from the switch to remote peer in order to slow down the traffic (if the peer
supports flow control).
In the other direction, when the switch receives a pause frame on a specific port from peer
equipment, the switch stops the packet transmission on that port until receives again a pause frame
with resume transmission command.
Flow control to be fully effective (no packets lost inside the network) requires that all devices in the
end-to-end path support flow control.
The flow control function is supported only when the capability is full duplex.
The flow control setting on the switch ports linked to user Ethernet ports must be consistent with the
setting on the user ports.
Flow control is supported on MSS-1c, on 1 port, in full duplex asymmetric Tx mode, meaning that the
switch will be able to transmit PAUSE frames, but will ignore received PAUSE frames.
Flow control is not supported on MPR-e.
9.5.2 Asymmetric Flow control This features on switch port based, allows of enable the pause frame only in transmission or receiver
side.
In the first case the switch can generate pause frame toward peer but is not able to stop
transmission traffic when receives a pause from peer.
45
In the second case, asymmetric receive flow control enabled, the switch when receives a pause
frame stops the transmission but is not able to transmit pause frame toward the peer.
The asymmetric flow control setting on the switch ports linked to user Ethernet ports must be
consistent with the setting on the user ports.
9.5.3 802.1Q VLAN management The port-based VLAN feature allows of partition the switch ports into virtual private domains.
According to the type of site configuration and cross-connections setting this feature is properly
managed by the software. For example, if all traffic from one Ethernet port must be forwarded only
in one radio direction is good to enable the traffic exchange only between these ports.
The IEEE 802.1Q tag VLAN feature can be enabled including between the other the stripping or
adding of the TAG and VLAN lookups in addition to MAC lookups (this feature between the other can
be useful for re-route TMN traffic to the controller).
The IEEE 802.1Q tag VLAN feature can be enabled or disabled (be transparent for the VLAN) including
between the other the stripping or adding of the TAG and VLAN lookups in addition to MAC lookups
(this feature can be useful to logically break a physical LAN into a few smaller logical LAN and to
prevent data to flow between the sub-LAN), dropping NON-VLAN Frames.
9.5.4 Link Aggregation (IEEE 802.3ad) Link Aggregation allows one or more physical links to be aggregated together to form a Link
Aggregation Group, such that a MAC Client (CES, VLAN Management, etc.) can treat the Link
Aggregation Group as if it is a single link.
Link Aggregation provides the following:
• Increased bandwidth: The capacity of multiple links is combined into one logical link
• Link protection: The failure or replacement of a single link within a Link Aggregation Group
does not cause failure from the perspective of a MAC Client.
• Load sharing: MAC Client traffic may be distributed across multiple links.
• Automatic configuration: Link Aggregation Groups are automatically configured and
individual links are automatically allocated to those groups relying on the Link Aggregation
Protocol.
46
• Static configuration: Link Aggregation Groups are statically configured by the operator.
Link aggregation is not currently supported on MSS-1c.
9.6 Ethernet OAM (IEEE 802.3ag) Ethernet OAM is a set of procedures for maintenance and troubleshooting of point-to-point and
multi-point Ethernet Virtual Connections that span one or more links. It is end-to-end within an
Ethernet network. The following figure shows a network comprising of multiple domains within the
metro network.
The customer subscribes to the services of a provider, who in turn subscribes to the services of two
operators. Every domain has its own NMS. There are two planes. “Vertical” plane in red shows the
OAM entities across different domains. “Horizontal” plane in blue has various OAM entities (MEPs
and MIPs) within a domain. The following figures show the cross-section across the vertical OAM
plane and the horizontal OAM plane respectively. The vertical plane figure shows a single monitored
path for each administrative domain; the horizontal plane figure shows two monitored paths for the
same administrative domain.
Customer domain
Provider domain
Operator 1domain Operator 2
domain
Customer domain
Provider domain
Operator 1domain Operator 2
domain
47
Levels
-
+
CustomerEquipment
CustomerEquipment
Operator ABridges
Operator BBridges
1 2 3 4 5 6 7 8 9
ETH
Maintenance End PointMaintenance Intermediate Point
Customer Level
Provider Level
Operator Level
ETH Section or
SRVServer Layer
MEP
MIP
Vertical plane cross-section
Horizontal plane cross-section
MIP1 MIP2 MIP3 MIP4 MIP5 MIP6 MIP7 MIP8
MIP9
MIP10
MIP11
MIP12
MEP1
MEP2
MEP3
MEP4
Bridge
PortMIP
MEP
48
Ethernet OAM provides the following tools:
Ethernet OAM will be supported on MSS-1c in future release.
49
9.7 Ethernet Ring Protection (ITU-T G.8032v2)
ERP allows a simple, Carrier Grade and reliable packet protection in ring topologies. It is applicable to
Full Microwave Rings only.
ITU-T G.8032v2 ERP filled the gap in Carrier grade Ring protection schema. (x)STP in fact has been
developed for LAN environments and it is not employed anymore in new network deployments for
its lack of determinism (depending on the position of root bridge) and scalability (BPDU needs to be
processed in each node, MSTP is complex to operate, Per-VLAN STP is not standardized and scalable)
in Carrier networks.
With reference to the following network scenario:
the following specifications apply:
• The ring is implemented by east and west facing radio directions
• Traffic can follow on both ring directions: Clockwise direction & Counter-clockwise direction
• Protection is triggered by physical criteria (no protocol intervention)
• Protection is based on R-APS messages sent on both sides of the ring by the nodes detecting
the failure. Traffic is redirected by each node of the ring locally, ensuring parallel processing
to speed up protection time.
• G.8032v2 algorithm operates on VLAN, regardless the type of traffic transported: TDM
(TDM2TDM and TDM2ETH) and Eth (Multiple CoS and services) traffic types can be protected
50
• Traffic flows (any type/priority) can be allocated on both ring directions to exploit the
maximum ring bandwidth in normal conditions for best effort traffic and to limit packet delay
when traffic enters from different points of the ring.
• G.8032v2 is supported on both MSS1/MSS4/MSS8
• Synchronization is managed through SSM messages (Synchronous Ethernet).
• Multichannel LAG L1 configuration can be supported inside the ring, with optionally error
error free adaptive modulation configured.
9500MPR does support ITU-T G.8032v2 in mixed configuration as well, meaning that some links can
be microwave and some links can use fiber. Here below few options available, with 8032v2 ring
implemented by ALU devices.
51
9.8 Other features
• Port Segregation: all traffic received/transmitted from one user Ethernet port or radio
direction cannot be exchanged with specific user Ethernet ports/radio directions
• Per flow policer: ingress rate limiter per VLAN, dropping the traffic exceeding a given CIR
value
• Per Cos policer: ingress rate limiter per p bits value (i.e. possibility to define a thresholds
above which the traffic a given pbit value or a given set of pbits values is dropped)
• Broadcast storm control: ingress rate limiter on broadcast traffic
• Multicast storm control: ingress rate limiter on multicast traffic
• MAC address access control list: only packet with SA inside a given list are transmitted
towards the radio
These features are not supported by MPR-e.
9.8.1 Stacked VLAN (Q-in-Q): 802.1ad
The switch supports double tagging according to 802.1ad, in particular:
• adding a service VLAN on the ingress traffic
• pbits value of service VLAN is a)user configurable b)same value of customer VLAN.
The EtherTypes supported are:
• EtherType 0x8100
• EtherType 0x9100
• EtherType 0x88A8
9.8.2 VLAN swap
Every incoming frames on a given user having VLANID xxx is remarked with VLANID yyy without
changing the priority (.1p bits).
This feature is not supported by MSS-1c and MPR-e
52
9.9 Ethernet QoS
The Ethernet switch provides a Quality of Service mechanism to control all streams. If by CT/NMS the
QoS is disabled all traffic inside the switch has the same priority, this means that for each switch port
there is only one queue (FIFO) therefore the first packet that arrives is the first that is transmitted.
9.9.1 Traffic priority
In the switch the QoS assigns the priority for each packet according to information in:
• Port-based: the same priority is assigned to each frame arriving at the given ingress port;
• IEEE std 802.1p: the packet is examined for the presence of a valid 802.1P user-priority Tag. If the
tag is present the correspondent priority is assigned to the packet;
• MAC based: the MAC destination address and VLAN ID are used to determine the priority for
each packet;
• DiffServ: each packet is classified based on DSCP field in the IP header to determine the priority;
By CT/NMS the priority can be chosen between 802.1p or DiffServ for each Network Element.
9.9.2 IEEE 802.1P QoS configuration
When 802.1p QoS mechanism is adopted the reference is the standard "IEEE 802.1D-2004 Annex G.
User priorities and traffic classes” that defines 7 traffic types and the corresponding user priority
values.
By CT/NMS is possible to configure the mapping 802.1p value to queue inside the switch (except for
MSS-1c).
When an incoming packet is not 802.1p it is assigned to the lowest priority queue.
9.9.3 DiffServ QoS configuration
When DiffServ QoS mechanism is adopted the classification uses the DS field of the IP packet header.
By CT/NMS is possible to configure the mapping DS field value to queue inside the switch (except for
MSS-1c). When an incoming packet has not DiffServ valid value it is assigned to the lowest priority
queue. IPv6 TOS classification is supported as well.
9.9.4 Congestion management
In case of traffic congestion is possible to choose between Random Early Detection (RED) or tail drop
algorithm before the congestion becomes excessive.
53
9.9.5 Quality of Service
Quality of service of CORE card: The Quality of Service feature of the Ethernet switch provides eight
internal queues for each port to support eight different class of service (COS). For each egress port
according to the method of QoS classification configured in the switch, the packets are assigned to
specific queue.
High priority traffic is served starting from Queue 8 to 6, while the remaining five queues are shared
by all generic Ethernet flows according the default and fixed classification mechanism configured by
CT/NMS.
In MSS-1c, classification services is slightly different to stick with specific requirements of the tail.
L2 switch in MSS-1c provides 4 internal queues per port
All TDM flows are assigned to highest egress priority queue (Q4)
Ethernet flows are assigned based on 802.1p or Diffserv information.
For MPR-e , the 3 first queues are dedicated to TDM2TDM, TDM2ETH and TMN traffic. TDM2TDM
and TDM2ETH traffic management will be supported in future release.
5 next queues are dedicated to Ethernet traffic.
For MPR-e, the Ethernet queues can be configured in HQP (starting from queue#5) in strict priority
algorithm to guaranty real time transport such as VoIP
#3
#2DWRR
MPR QoS
HPQTDM #4
ETHERNET
ETHERNET
ETHERNET
Scheduler type
Service type
#1
54
Two types of scheduler algorithms are possible:
• Deficit Weighted Round Robin (DWRR); the weights determine the number of blocks (not the
number of packets) that each queue can send at each algorithm round.
• Strict Priority (SP) or High Queue Preempt (HQP); guarantee that when the queue with higher
priority is not empty, it is immediately served. The primary purpose of the strict priority
scheduler is to provide lower latency service to the higher CoS classes of traffic.
Classification
VLAN&MAC
VLAN&MAC
VLAN&MAC
1p/Diffserv
Scheduler type
Service type
MPR QoS
HPQ
TDM
TDM2ETH
TMN
#8
#7
#6
ETHERNET
ETHERNET
ETHERNET
ETHERNET
ETHERNET
1p/Diffserv
1p/Diffserv
1p/Diffserv
1p/Diffserv #1
#5
#4
#3
#2
HPQ/DWRR
55
10 Radio Configurations
The following configurations are available for each radio path.
1+0
In this configuration the radio chain consists of:
� One Radio Outdoor Unit (MPT)
� One Antenna
� One MPT Access Card
1+1
In this configuration the radio chain consists of:
� Two Radio Outdoor Units ( MPT)
� One or two antennas
� One or two MPT Access Cards
�
Following options are available for protected configuration:
� Hot Stand-by (with or w/o coupler)
� Frequency Diversity
� Polarization Diversity
1+1 Hot Standby
This method offers protection against HW failures providing two independent TX/RX chains. In
(1+1)HSby one transmitter is working, while the other one is in stand-by; both receivers are active
and the best ODU source is selected.
In case of 1+1 Hot Stand-by on single antenna, both Radio Units are connected to a coupler, balanced
or un-balanced.
(1+0)
ODU 300/MPT
Modem/AWY/ MPT
Access Card
56
Alternatively, in case of 1+1 Hot Stand-by Space Diversity, each Radio Unit is connected to an
individual antenna.
.
1+1 Frequency Diversity/Polarisation DIversity
This method offers protection against selective and temporary link quality degradation.
(1+1)Hsby On two Antennas
Main
Hsby
Modem/MPT Access Card
(1+1)Hsby On Single Antenna
Main
Hsby
Modem/MPT Access Card
57
In (1+1) Frequency Diversity, both radio paths are active in parallel using different frequencies; this
method, based on memory buffer that guarantees the bit to bit alignment, can offer error free
protection against fading (via a hitless switch) up to 100dB/sec.
Both two antennas and single antenna (dual polarized) mounting arrangements are available.
(However, with FD, the usual arrangement is one antenna SP.)
(1+1) Polarization Diversity adopts the same concepts of FD, but in this case the same RF signal is
transmitted on two different polarizations (H/V) by means of a single double polarized antenna.
Adjacent Channel Alternate Polarised (ACAP), Adjacent Channel Co Polarised (ACCP) and Co-Channel
Dual Polarisation (CCDP) operations are supported
10.1 Antenna Mount
Direct-Mounted Radio Unit
The Radio Unit is attached to its antenna by a direct-mount collar, which includes a built-in rotator
for selection of vertical or horizontal polarization.
A full range of direct-mount antennas is offered with diameters from 0.3m to 1.8m. As an aid to
antenna alignment, the ODU includes receive signal level (RSL) access
(1+1) Frequency/Polarisation Diversity On two Antennas
Main
Hsby
Modem/MPT Access Card
F1/H1
F2/H2
58
For single antenna protected, frequency diversity and 2+0 operation, a direct-mount antenna coupler
for two ODU is available.
11 Power Supply
9500 MPR operates with nominal a -48 VDC power supply (positive grounded)in a voltage range of –
40.5 to –57.6V DC.
The DC power supply must be UL or IEC compliant for a -48V DC SELV output. The MSS-8 has the +Ve
pin on its DC power supply connector fastened directly to the shelf so must be used with a -48V DC
power supply which has a +Ve earth; the power supply earth conductor is the +Ve supply to the
radio. There must be no switching or disconnecting devices in this earth conductor between the DC
power supply and the point of connection to the radio.
MSS4/8 shelves are protected against polarity inversion, i.e. in case of inversion of "+" and "-" poles.
In this case, simply the equipment does not switch on and there are no damages in the equipment.
Power Distribution
The system receives the Battery input through 2 power connectors mounted on the shelf (MSS-8
shelf only) and connected directly to the Backplane. MSS-4 and MSS-1 shelf have 1 power connector.
Each board receives the Battery input (via Backplane) and provides adaptation to the customer
central power bus.
The input voltage range is from –40.5 to –57.6V DC. Nominal Voltage is –48V DC - Positive grounded.
Power Protection
59
Two different topics have to be considered:
1. DC/DC converter protection :
9500 MPR does adopt a distributed power supply architecture, meaning that each card has its own
DC/DC converter. Consequently no single point of failure is present and powering is fully protected.
