Nokia EGPRS3 Details

5/12/2018 Nokia EGPRS3 Details - slidepdf.com

http://slidepdf.com/reader/full/nokia-egprs3-details 1/55

1 © NOKIA EGPRS radio_v_2.PPT/ 23.3.2001 Company Confidential

EGPRS Radio interface

Packet data logical channels

Data transfer principles

RLC Radio block structure

Multislot MS MCS selection

Multiplexing GPRS/EGPRS

Link adaptation




Packet data logical channels 1/3

PBCHPacket Broadcast Channels

PCCCH

Packet Common Control Chan

PDCCHPacket Dedicated Control Cha

PTCHPacket Traffic Channels




Packet data logical channels 2/6

PBCCH available in BSS S10 GPRS and EGPRS will not impact on AGCH and PCH

Increased data rates will lead to heavier signalling demandfor given traffic occupancy and applications

Note - PBCCH traffic is not carried on TRXSIG Note - PBCCH must be on the same TRX as the BCCH

GMSK is used on packet control channels. 8-PSKmodulation is used only on the packet traffic data channelPDTCH.

PBCCH/PCCCH

PBCCH PPCH PDTCH PRACH

BCCH/CCCH

Common Channels

PAGCH




Packet Broadcast control channel

PBCCH broadcast packet data specific System Information.If PBCCH is not allocated, the packet data specific system information is

broadcast on BCCH.

A broadcast control channel is a point-to-multipoint uni-directional control channel, from the fixed sub-system to theMobile Stations. Broadcast control channels are physicallysub-divided into Broadcast Control Channel (BCCH),Packet Broadcast Control Channel (PBCCH) and CompactPacket Broadcast Control Channel (CPBCCH).




Packet data Common Control logicalchannels

PAGCH is used in the packet transfer establishment phase tosend resource assignment to an MS prior to packet transfer.

PNCH is used to send a PTM-M (Point To Multipoint -Multicast) notification to a group of MSs prior to a PTM-Mpacket transfer. ( NOT IMPLEMENTED in S10 )

PRACH is used by MS to initiate uplink transfer for sendingdata or signalling information.

PPCH is used to page an MS prior to downlink packet transfer.




Paging Coordination

PPCH is used to page an MS prior to downlink packet transfer.PPCH CPPCH use paging groups in order to allow usage of DRX mode. PPCH canbe used for paging of both circuit switched and packet data services. The paging for circuit switched services on PPCH is applicable for class A and B GPRS MSs inNetwork operation mode I, see 3GPP TS 23.060

Mode Circuit Paging Channel GPRS Paging Channel Paging co-ordination

Packet Paging Channel Packet Paging Channel

I CCCH Paging Channel CCCH Paging Channel Yes

Packet Data Channel Not Applicable

II CCCH Paging Channel CCCH Paging Channel No

III CCCH Paging Channel Packet Paging Channel No

CCCH Paging Channel CCCH Paging Channel

Paging co-ordination = CS pages can be sent through SGSN and via PPCH




Packet data dedicated channels

PACCH conveys signalling information related to a given MS.e.g. acknowledgements, power control, resource assignment and reassignmentmessages.. The PACCH shares resources with PDTCHs, that are currentlyassigned to one MS. Additionally, an MS that is currently involved in packet transfer,can be paged for circuit switched services on PACCH.

PTCCH is used to transmit random access burst to allow

estimation of the timing advance for MS's in packet transfer mode.

PDTCH is a channel allocated for data transfer. It is temporarilydedicated to one MS or to a group of MSs in the PTM-M case. In the multislotoperation, one MS may use multiple PDTCHs in parallel for individual packet

transfer.




Mapping of PBCCH/PCCCH DL

B0 B1 B2 T B3 B4 B5 X B6 B7 B8 T B9 B10 B11 X

X = Idle frame

T = Frame used for PTCCH

B0 - B11 = Radio blocks

A physical channel allocated to carry packet logical channels is called a packetswitched channel (PDCH). A PDCH shall carry packet logical channels only.Packet switched logical channels are mapped dynamically onto a 52-multiframe. The 52-multiframe consists of 12 blocks of 4 consecutive frames, 2idle frames and 2 frames used for the PTCCH, as shown in figure. PCU handlesradio blocks 0 ± 11. BTS handles the rest of the radio blocks.




y BS_PBCCH_BLKS, number of PBCCH blocks per multiframe.Depends on the number of PSIs and repeat period.

y BS_PAG_BLKS_RES, number of blocks in addition toBS_PBCCH_BLKS, where paging shall not occur.

y BS_PCC_CHANS number of physical channels carryingPCCCHs including the physical channel carrying the PBCCH.

FIXED to 1

For example:

BS_PBCCH_BLKS = 3 (number of blocks allocated to PBCCH per multiframe)

BS_PAG_BLKS_RES = 4 (number of blocks allocated to PAGCH per multiframe)

Empty blocks can be used for packet assignment, packet paging, packet data or associated signalling. PAGCHblocks can be used for packet assignment, packet data or associated signalling

PBCCH PAGCH PBCCH PAGCH PBCCH PAGCH PAGCH

BL 0 BL 1 BL 2 BL 3 BL 4 BL 5 BL 6 BL 7 BL 8 BL 9 BL 10 BL 11

Mapping of PBCCH/PCCCH DL (2/3)




Data Transfer - Temporary Block Flow

A Temporary Block Flow (TBF): is a one-way session for packet data transfer between MS and

BSC (PCU)

uses either uplink or downlink but not both (except for

associated signalling)

can use one or more TSLs

Compare with circuit-switched - normally one connection usesboth the uplink and the downlink timeslot(s) for traffic.

In two-way data transfer, uplink and downlink data are sent in

separate TBFs - as below.

BSCBSC

Uplink TBF (+ PACCH for downlink TBF)

Downlink TBF (+ PACCH for uplink TBF)




Establishing a TBF and Sending Data-UL

Channel request - RACH

Immediate Assignment - AGCH

Data - PDTCH

Signalling + Ack/Nack - PACCH

BTSBTS

Final Data - PDTCH

Final Ack/Nack - PACCH

Packet control Ack - PACCH




Establishing a TBF and Sending Data -DL

Immediate Assignment - AGCH

Paging - PCH

UL TBF for MS location

Packet control Ack - PACCH (for TA)

Packet polling - PACCH)

Packet downlink assignment - PACCH

Data - PDTCH/ Signalling -PACCH

Ack/Nack - PACCH

BTSBTS




Radio Resource operating modes

Packet idle mode No Temporary Block Flow exists MS listens to the PBCCH and to the paging sub-channel for the paging

group the MS belongs to in idle mode. If PCCCH is not present in the cell,the mobile station listens to the BCCH and to the relevant paging sub-channels

Packet transfer mode MS is allocated radio resource providing a Temporary Block Flow on one or

more physical channels When selecting a new cell, mobile station leaves the packet transfer mode,

enters the packet idle mode where it switches to the new cell, read thesystem information and may then resume to packet transfer mode in thenew cell

Dual transfer mode MS has an ongoing RR connection and is allocated radio resource

providing a Temporary Block Flow on one or more physical channels While in dual transfer mode the MS performs all the tasks of dedicated

mode



14 © NOKIA EGPRS radio_v_2.PPT/ 23.3.2001Company Confidential

GPRS Mobility Management States

Idle

Standby

Ready

PacketTX/RXSTANDBY

Timer Expiry

GPRS Attach/ Detach

READYTimer Expiry

MS location known toRA level.MS is capable of beingpaged for data

MS locationnot known.Subscriber is notreachable by theGPRS NW.

MS location known tocell level.MS is transmitting or has

just been transmitting.




RR operating modes and MM Statescorrespondence

Non-DTM capable MS:

RR BSSPacket

transfer modeMeasurement

report receptionNo state No state

RR MSPacket

transfer modePacket idle mode

Packetidle mode

MM (NSS

and MS)Ready Standby

DTM capable MS:

RR BSSMeasurement

report receptionNo state No state

RR MS

Dualtransfer

mode

Dedicatedmode

Packettransfer

mode CS idle and packet idle

Dedicatedmode CS idle and

packet idleGMM (NSS

and MS)Ready Standby




EGPRS (Enhanced General Packet Radio Service) is an improved air interface for GPRS packet service. The peak throughput of EGPRS is 2.96 times the peak throughput of GPRS.

The figure below shows the (E)GPRS protocol stack when carrying IP data.Radio performance depends mostly on the RLC/MAC and physical layer of the radio interface (Um).

(E)GPRS Transmission plane

Scope of 3GPP TS23.060

Relay

Network

Service

GTP

Application

IP / X.25

SNDCP

LLC

RLC

MAC

GSM RF

SNDCP

LLC

BSSGP

L1bis

RLC

MAC

GSM RF

BSSGP

L1bis

Relay

L2

L1

IP

L2

L1

IP

GTP

IP / X.25

Um Gb Gn GiMS BSS SGSN GGSN

Network

Service

UDP /

TCP

UDP /

TCP

Scope of TS 03.64



Logical Link Control (LLC) Layer

LLC

SNDCP

IP

TCP/UDP

APP

RLC

MAC

GSM RF

Reliable logical connection between SGSN and MS

Independent of underlying radio interface protocols

Control Address

FCSInformation

LLC Frame

1 1-3 1-1520 3 Octets



Radio Link Control (RLC)/ MediumAccess Control (MAC) Layers

RLC

Reliable transmission of data across air interface

Segmentation/de-segmentation of data from/toLLC layer

MAC

Control of MS access to common air-interfacemedium

Flagging of PDTCH/PACCH occupancy

LLC

SNDCP

IP

TCP/UDP

APP

RLC

MAC

GSM RF




RLC/MAC

Implemented in the PCU (Packet Control Unit, part of BSC).

TBF (Temporary Block Flow) is a one-directional data flowbetween two RLC/MAC entities (PCU and MS).

TBF is established when new data arrives to RLC.

TBF is released when all the data in RLC has been sent. Acknowledged RLC mode:

ARQ, retransmissions and IR. RLC/MAC can guarantee error-free and in-order delivery of

upper layer blocks..

Unacknowledged RLC mode: No ARQ, retransmissions or IR. Corrupted blocks are lost or delivered in error.




data to be transmitted

RLC: Segmentation & ARQ

Segmentation makes RLC blocksfrom LLC frames.

Size of RLC block matches theMCS.

LLC frames can be small or large. Large LLC frames takemany RLC blocks.

RLC blocks can contain LLCframe boundaries (even multiplesmall LLC frames).

ARQ (Automatic Repeat reQuest)takes care of backward error correction, i.e. requesting andperforming retransmissions of incorrectly received blocks.

Used in both GPRS andEGPRS when in acknowledgedRLC mode.

Multiple blocks are

acknowledged at the same time.

1 20 43

? 0? 1

? 2?

3

4

0

2

1 20 43

transmitter

receiver

time

all data correctly received




RLC: Retransmissions

originaltransmission

normalretransmission

retrans withresegmentation

MCS-1 MCS-1

MCS-2 MCS-2

MCS-3 MCS-3

MCS-4 MCS-4 2 * MCS-1

MCS-5 or 7 MCS-5 or 7 2 * MCS-2MCS-6 or 9 MCS-6 or 9 2 * MCS-3

MCS-8 MCS-8 MCS-6 (pad)2 * MCS-3 (pad)

In GPRS, retransmissions arealways in the same CS as theoriginal transmission, whichcauses problems if too high a CShas been selected.

To help this, EGPRS allows aretransmission to be performedin a lower MCS than original.

MCSs are divided in families: C: MCS-1, MCS-4 B: MCS-2, MCS-5, MCS-7 A: MCS-3, MCS-6, MCS-8, MCS-9

With normal (not resegmented) retransmissions the structure and content of the RLC block is identical to the original transmissions. Therefore IRcombining is possible.

In resegmentation, an RLC block is split in two halves. Padding meansadding dummy bits to an MCS-8 block so it looks like an MCS-6 block. IRcombining is not possible.




MAC: Scheduling

(E)GPRS timeslots can be sharedby multiple MSs.

The radio resource is allocated ona radio block (4 bursts, 20 ms)resolution by the network.

In the downlink, all the MSs listencontinuously. TFI (Temporary FlowIdentifier) is used to address oneMS.

In the uplink, there are differentresource allocation modes: fixed,dynamic and extended dynamic

allocation.

Fixed allocation A large number of (not

necessarily continuous) radioblock periods are allocated to aMS in advance using a controlmessage. (Not used in Nokiaimplementation.)

Dynamic allocation NW sends USFs (Uplink Status

Flags) in each downlink block tosignal which MS is allowed totransmit during the next uplinkblock period.

The USF may be for a differentMS than the actual data in thedownlink radio block. Problemswith power control and smartantennas.

Extended dynamic allocation

multislot




Throughput

Main performance indicator of GPRS & EGPRS.

Throughput is defined as the number of user data bitsdelivered per second (usually per time slot).

The definition of user data bit excludes the following overhead

from the gross rates: physical layer overhead due to channel coding, RLC/MAC layer overhead due to header information, overhead due to retransmissions.




EGPRS Modulation and CodingSchemes

EGPRS has nine basic codingschemes, MCS-1...9.

In general, a higher coding schemehas higher coding rate, andconsequently higher peakthroughput, but it also tolerates less

noise or interference. The figure shows throughput vs. C/I

of EGPRS coding schemes inTU50iFH, without incrementalredundancy.

The basic unit of transmission is

radio block (= 4 bursts = 20 ms onaverage), which contains one or two RLC blocks.

0

10

20

30

40

50

60

0 5 10 15 20 25 30

MCS-1

MCS-2

MCS-3

MCS-4

MCS-5

MCS-6

MCS-7

MCS-8MCS-9




Data treatment principle in RF layer

User data

"Additional info" that does not require extraprotection

Header part, robust coding for securetransmission




EGPRS Channel Coding

EGPRS channel coding consistsof separate data and header coding, as shown in the figurefor MCS-9 downlink.

Coding of data part: Data part includes user

data, two bits from RLCheader, BCS (block check sequence)and tail bits.

Coded using 1/3 convolutional code. Punctured with a selectable puncturing scheme

(P1, P2 or P3). Two separate data parts for MCS-7...9.

Header part: Includes RLC/MAC header information and

information on the coding of the data part (likeused puncturing scheme).

Convulutional coding + puncturing.

USF

encoded USF P2 P3

P1 P2 P3

puncturing puncturing

puncturing

1st burst 2nd burst 3rd burst 4th burst

1/3 tailbitingconvolutional coding

blockcoding

P1

header FBI+E data 2 BCS tail

1/3 convolutionalcoding

FBI+E data 1 BCS tail

1/3 convolutionalcoding

mother code

mother code

protectedheader




GPRS & EGPRS Coding Schemes

codingscheme

modulation RLC blks /radio blk

FECcode rate

user bits /20 ms

bit rate(bps)

CS-1 1 0.45 160 8,000

CS-2 1 0.65 240 12,000

CS-3 1 0.75 288 14,400

GPRS

CS-4 1 n/a 400 20,000

MCS-1 1 0.53 176 8,800

MCS-2 1 0.66 224 11,200

MCS-3 1 0.85 296 14,800

MCS-4

GMSK

1 1.00 352 17,600

MCS-5 1 0.38 448 22,400

MCS-6 1 0.49 592 29,600MCS-7 2 0.76 448+448 44,800

MCS-8 2 0.92 544+544 54,400

EGPRS

MCS-9

8-PSK

2 1.00 592+592 59,200




EGPRS MCS´s

37 octets 37 octets 37 octets37 octets

MCS-3

MCS-6

Family A

MCS-9


MCS-2

MCS-5

MCS-7

Family B

22 octets22 octets

MCS-1

MCS-4

Family C

34+3 octets34+3 octets

MCS-3

MCS-6

Family A

padding

MCS-8


The MCSs are divided into different families

A,B and C.

Each family has a different basic unit of payload: 37 (and 34), 28 and 22 octetsrespectively.

Different code rates within a family areachieved by transmitting a different number of payload units within one Radio Block.

For families A and B, 1 or 2 or 4 payload unitsare transmittes, for family C, only 1 or 2payload units are transmitted

When 4 payload units are transmitted (MCS 7,

MSC-8 and MCS-9), these are splitted into twoseparate RLC blocks (with separate sequenceBSN numbers and BCS, Block CheckSequences)

The blocks are interleaved over two burstsonly, for MCS-8 and MCS-9.

For MCS-7 the blocks are interleaved over




Interleaving over 2 bursts(header: 4bursts)

Decreasingredundancy

Addingredundancy

MCS-9 coding and puncturing

P2 P3P1 P2

puncturing puncturing

1836 bits

USF RLC/MACHdr.

36 bits

Rate 1/3 convolutional coding

135 bits

612 bits

612 bits124 bits36 bitsSB = 8

1392 bits

45 bits

Data = 592 bits BCS TB

612 bits

612 bits 612 bits

1836 bits

Rate 1/3 convolutional coding

EFBIData = 592 bits BCS TBEFBI

612 bits 612 bits 612 bits

P3 P1

3 bits

HCS

puncturing

Ref: TS 03.64

1 1 12 6

Data rate: skbms

/2,5920

5922!

v

Robust codingfor header

Normal burst:2x58x3 bits




BP: 15/26 ms BP: 15/26 ms BP: 15/26 ms BP: 15/26 ms

skbms

/6,6920

1392!

20 ms




Coding parameters

EGPRS modulation and coding schemes:

Scheme Code rate Header Code rate

Modulation RLC blocksper Radio

Block(20ms)

Raw Datawithin one

Radio Block

Family BCS Tailpayload

HCS Data ratekb/s

MCS-9 1.0 0.36 2 2x592 A 59.2

MCS-8 0.92 0.36 2 2x544 A 54.4

MCS-7 0.76 0.36 2 2x448 B

2x12 2x6

44.8

MCS-6 0.49 1/3 1 592544+48

A 29.627.2

MCS-5 0.37 1/3

8PSK

1 448 B 22.4

MCS-4 1.0 0.53 1 352 C 17.6

MCS-3 0.80 0.53 1 296272+24

A 14.813.6

MCS-2 0.66 0.53 1 224 B 11.2

MCS-1 0.53 0.53

GMSK

1 176 C

12 6

8

8.8

NOTE: the italic captions indicate the padding.

Ref: TS 03.64




EGPRS RLC/MAC header

Bit8 7 6 5 4 3 2 1 Octet

TFI RRBP ES/P USF 1

BSN1 PR TFI 2BSN1 3

BSN2 BSN1 4

CPS BSN2 5

Bit8 7 6 5 4 3 2 1 Octet

TFI Countdown Value SI R 1

BSN1 TFI 2

BSN2 BSN1 3

BSN2 4

Spare PI RSB CPS 5

Spare 6

Downlink:

Uplink:

Ref: TS 04.60

Three header types for EGPRS

Example: Header type 1 (header for MCS-7, MCS-8 and MCS-9)




Multiple Mobiles and UplinkTransmission

USF =1

USF = 2

USF =3

USF = 3

MSs

BTS

RLC Data Block

Mobile transmissions controlled by USF (Uplink State Flag)sent on DL

Mobile with correct USF will transmit in following block




Multiple Mobiles and DownlinkTransmission

TFI2

TFI5

TFI3

TFI2

MSs

BTS

TFI value included in RLC block header - indicateswith which TBF the RLC block is associated

RLC Data Block




MS classes for multislot capability Rx: Rx describes the maximum number of receive timeslots that the MS can

use per TDMA frame. The MS must be able to support all integer values of receive TS from 0 to Rx (depending on the services supported by the MS). Thereceive TS need not be contiguous. For type 1 MS, the receive TS shall beallocated within window of size Rx, and no transmit TS shall occur betweenreceive TS within a TDMA frame.

Tx: Tx describes the maximum number of transmit timeslots that the MS canuse per TDMA frame. The MS must be able to support all integer values of transmit TS from 0 to Tx (depending on the services supported by the MS). Thetransmit TS need not be contiguous. For type 1 MS, the transmit TS shall beallocated within window of size Tx, and no receive TS shall occur betweentransmit TS within a TDMA frame.

Sum: Sum is the total number of uplink and downlink TS that can actually beused by the MS per TDMA frame. The MS must be able to support allcombinations of integer values of Rx and Tx TS where 1 <= Rx + Tx <= Sum(depending on the services supported by the MS). Sum is not applicable to allclasses.




Multislot Classes

29 multislot classes specified: e.g.Multislot

classMaximum number of slots Minimum number of slots Type

Rx Tx Sum Tta Ttb Tra Trb

1 1 1 2 3 2 4 2 12 2 1 3 3 2 3 1 13 2 2 3 3 2 3 1 14 3 1 4 3 1 3 1 15 2 2 4 3 1 3 1 16 3 2 4 3 1 3 1 17 3 3 4 3 1 3 1 18 4 1 5 3 1 2 1 19 3 2 5 3 1 2 1 1

10 4 2 5 3 1 2 1 111 4 3 5 3 1 2 1 112 4 4 5 2 1 2 1 113 3 3 NA NA a) 3 a) 214 4 4 NA NA a) 3 a) 215 5 5 NA NA a) 3 a) 216 6 6 NA NA a) 2 a) 2

17 7 7

NA NA a) 1 0 218 8 8 NA NA 0 0 0 219 6 2 NA 3 b) 2 c) 120 6 3 NA 3 b) 2 c) 121 6 4 NA 3 b) 2 c) 122 6 4 NA 2 b) 2 c) 123 6 6 NA 2 b) 2 c) 124 8 2 NA 3 b) 2 c) 125 8 3 NA 3 b) 2 c) 126 8 4 NA 3 b) 2 c) 127 8 4 NA 2 b) 2 c) 1

28 8 6 NA 2 b) 2 c) 129 8 8 NA 2 b) 2 c) 1

Tta: Tta relates to the time needed for the MS to perform adjacent cell signal level

measurement and get ready to transmit.

Ttb: Ttb relates to the time needed for the MS to get ready to transmit. This minimum

requirement will only be used when adjacent cell power measurements are not required by the

service selected.

Tra: Tra relates to the time needed for the MS to perform adjacent cell signal levelmeasurement and get ready to receive.

Trb: Trb relates to the time needed for the MS to get ready to receive. This minimum

requirement will only be used when adjacent cell power measurements are not required by the

service selected.




MS multislot example

Multislot class 5 MS in circuit switched configuration Five basic configurations of channels are possible

Rx

Tx

Rx=2

Tt=1 Tx=2

Rx

Tx

Rx=2

Tt=1

Tx<2

Rx

Tx

Rx<2 Tt>1 Tx=2

These five combinations can be repeatedat the six other positions that can be fitted

within the same TDMA frame

Rx

Tx

Rx=2 Tt>1

Tx<2

Tra>3

Tra=3 Tra=3

Tra=3

Rx

Tx

Rx<2 Tx=2 Tra>3

Tt=1

All possible timeslots used Downlink biased assymetry

Alternative downlink biased assymetry Uplink biased assymetry(not prohibited by multislot class)

Alternative uplink biased assymetry(not prohibited by multislot class)Ref: TS 05.02




Multiplexing GPRS/EGPRS

GPRS and EGPRS mobile stations can be multiplexed dynamically on thesame PDCH

Timeslot scheduling algorithm determines if there is a standard GPRS MSand EGPRS MS multiplexed on the timeslot, then at least one Radio Blockevery 360ms on the downlink must use GMSK coding scheme.

For MS synchronisation resons, if standard GPRS MS are multiplexed on the PDCH,

at least one Radio Block every 360 ms on the Downlink must use GMSK (i.e.standard GPRS or MCS-1to MCS-4) (GSM 04.60 v. 8.5.0)

When USF is sent to GPRS MS the downlink coding scheme used inmultiplexed timeslot shall be restricted to MCS1 to MCS4




Incremental Redundancy

Link Adaptation




Incremental Redundancy (1) IR is a physical layer performance enhancement for the

acknowledged RLC mode of EGPRS.

The basis for Incremental Redundancy (IR) is in the selective-reject- ARQ protocol of the RLC layer. The ARQ protocol takes care of requesting and retransmitting incorrectly received blocks.

By using the Backward Error Correction (BEC) procedures the

selective retransmission of unsuccessfully delivered RLC/MACblocks is obtained

IR improves the reception of retransmissions by combining theinformation in the original transmission (which failed) with thereceived additional information, thereby increasing the probability of correct reception.

The most important standardised feature of IncrementalRedundancy is that MS has mandatory IR combining in its receiver.IR has also been taken into account in the design of the codingschemes and block formats.




Incremental Redundancy (2)

The figure shows an example of IR transmission andcombining with different puncturing schemes for differenttransmission. The shown case corresponds to MCS-4 or MCS-9, where the basic code rate is 1/1.

original data

1/3 coded data

1st xmission

2nd xmission

3rd xmission

1st decoding attempt

2nd decoding attempt

3rd decoding attempt

r = 1/3

r = 1/2

r = 1/1

r = 1/1

r = 1/1

r = 1/1




0

10

20

30

40

50

60

0 5 10 15 20 25 30

MCS-9

MCS-9 IR

MCS-9 IR 20 ksv


MCS-9 with and without IR. Throughput vs. C/I in TU3 ideal FH.

Gain of IR is up to 11 dB in thiscase.

The dashed line shows IRperformance with a limited IRreceiver memory (20480 softvalues). The IR gain is now limitedto 8 dB.





Ideal MCS selection with andwithout IR.

Throughput vs. C/I in TU3 ideal FH.

Gain of IR is at most 2 dB.

Gains of IR and LA are not additive.

In reality, MCS selection is sub-optimal, so the real IR gain isbetween 2 and 8 dB.

In other words, IR has two benefits: easier MCS selection, performance enhancement.

0

10

20

30

40

50

60

0 5 10 15 20 25 30

ideal MCS

ideal MCS IR




Link adaptation-MCS Selection-

Link adaptation comprises an initial MCS selection and MCSadaptation

In the uplink, the entire MCS selection is controlled by oneparameter, and MS performs MCS selection according to

rules in 04.60. In the downlink, MCS selection is performed by the network.

This makes trying different MCSs easier and allows finer control in choosing the MCS for retransmissions. Alsospecial situations such as lack of IR memory can behandled separately on a block-by-block basis.

For initial transmissions: If IR memory is not full, use MCSpref_IR . If IR memory is full, use MCSpref_noIR .




Link adaptation-adaptation-

S10 Link adaptation algorithm:

is based on BEP mean and variance, reported by MS does not need TS specific BEP values

BEP mapping tables are based on extensive link levelsimulations (they may later be updated according to e.g.field measurements)

does not take BLER (ACK/NACK information) into account has 6 operator parameters which provide good and stable

control of the algorithm is simple to implement and do not require much computing

or additional memory




Link Adaptation: Introduction 1/3

The task of the link adaptation algorithm is to select the optimalMCS for each radio condition to maximise channel throughput.

Normally, LA adapts to path loss and shadowing but not fastfading. This corresponds to the "ideal LA" curves in link levelsimulations.

For very fast moving mobiles LA cannot adapt to shadowing,which will be handled by IR.

For very slow moving mobiles, the effect of fast fading cannotbe separated from the effect of shadowing, therefore LA willadapt to fast fading.

LA must take into account if IR combining is performed at thereceiver.

LA must take into account the effect of finite IR memory.

LA must also support the unacknowledged RLC mode




Link Adaptation: MCS selection

The default initial MCS value for RLC link adaptation isreceived from parameter database and it is an operator definable parameter

one default initial MCS value for available frequency bandsis used

MAC receives the initial MCS setting when MAC requestsfor an allocation

MAC informs RLC of initial MCS when TBF is created

EGPRS Link Adaptation is activated in downlink when RLCreceives EGPRS Packet Downlink ACK/NACK message onEGPRS acknowledge mode TBF

In unacknowledged mode Link Adaptation is used to makeMCS selection to downlink packets without packetretransmission

The initial MCS value for UL data transfer is received by MS in

PACKET UPLINK ASSIGNMENT or in PACKET TIMESLOT




Link Adaptation: RetransmissionControl

Acknowledged RLC mode uses selective packet confirmationwhere the receiving side confirms incoming packets.

RLC packets are BSN numbered within TBF in use, and theBSN numbers are used to identify packets not receivedcorrectly

Each RLC data block contains a block sequence number (BSN) field that is 7 bits in length

[04.60 / 9.1.4]

Receiving side sends confirmations either EGPRS PACKETDOWNLINK ACK/NACK or PACKET UPLINK ACK/NACKdepending on TBF transfer direction.

The EGPRS PACKET DOWNLINK ACK/NACK confirmation packet is

sent by the MS when requested by the network.

The PACKET UPLINK ACK/NACK confirmation packet is sent by thenetwork whenever the network decides to do so

In downlink TBF the network can query the MS to send aEGPRS Channel Quality Report IE




Changing the MCS of an RLC block MCSs are divided into three families: (MCS-1, MCS-4),

(MCS-2, MCS-5, MCS-7) and (MCS-3, MCS-6, MCS-8, MCS-9). Switching across families is never possible.

RLC blocks of MCS-6 and MCS-9 can be freely switched backand forth

the difference between the two RLC blocks is that in MCS-9 the mother

code is punctured more so that there is space in the radio interfaceblock for two RLC blocks

also IR combining is possible between these two MCSs.

Same applies to MCS-5 and MCS-7.

IR combining is impossible across all the other switches, which

are: MCS-4 -> MCS-1-split (original RLC block is splitted into two) MCS-5 or MCS-7 -> MCS-2-split (original RLC block is splitted into two) MCS-6 or MCS-9 -> MCS-3-split (original RLC block is splitted into two) MCS-8 -> MCS-6-pad -> MCS-3-pad-split (MSC-8 block can not be

directly transmitted by using lower MCS due to number of payload bitsis not an integer multiple of a lower MCS number of paylod bits,




Link Adaptation Algorithm

Standard has specified the BEP (Bit Error Probability)measurements to be done by the MS (in DL)

the BEP is done from the "soft bits" generted by the receiver (from the data bits)

soft bit means decision which includes information about thereliability of the decision (i.e. how sure the receiver is thatthe received bit is decided correctly) - BEP can becalculated from that certainty information

the certainty value is a characteristic of viterbi decoder usedin GSM reception





The LA procedure is based on static MCS selection tables inthe PCU. MCS selection can be divided in three classes:

1. Initial MCS to be used when entering the packet transfer mode

2. MCS selection for initial transmissions of each RLC blockin ACK mode

3. MCS to be used for retransmissions

4. Modulation selection

The algorithm is activated on dowlink whenever ameasurement report from MS is received. The algorithm is

activated on uplink whenever the channel coding command is

to be transmitted from network to the MS.




Link Adaptation Algorithm 1

Initial MCS to be used when entering the packet transfer mode

In this case, the MCS to be used is decided according to theoperator parameters:

y "Initial MCS coding scheme for the acknowledged mode"

y "Initial MCS coding scheme for unacknowledged mode"





MCS selection for initial transmissions of each RLC block in ACK mode

MCS selection is based on the MEAN_BEP and CV_BEP of the selected modulation.

The appropriate table below is consulted for MCS selectionfor GMSK and 8-PSKCV_BEP-class

MEAN_BEP-class

0 1 2 3 4 5 6 7

0 ± 3 1 1 1 1 1 1 1 1

4 2 2 1 1 1 1 1 1

5 2 2 2 1 1 1 1 1

6 2 2 2 2 2 2 1 1

7 ± 9 2 2 2 2 2 2 2 210 ± 19 3 3 3 3 3 3 3 3

20 ± 31 4 4 4 4 4 4 4 4

For GMSK

CV_BEP-classMEAN_BEP-class

0 1 2 3 4 5 6 7

0 ± 3 5 5 5 54 5 5 5 5

5 5 5 5 5

6 5 5 5 5

7 5 5 5 5

8 5 5 5 5

9 6 510 ± 16 6 6 6 6 6 6 6 6

17 ± 21 7 7 7 7 7 7 7 7

22 ± 25 8 8 8 8 8 8 8 826 ± 31 9 9 9 9 9 9 9 9

For 8-PSK





MCS selection downlink for retransmissions in ACK mode If modulation selection has selected to GMSK, GMSK will

also be used for retransmissions of 8-PSK blocks bysplitting the block (TBC)

For 8-PSK modulation, the table below shows the maximumMCS used for retransmissions.

CV_BEP-classMEAN_BEP-class

0 1 2 3 4 5 6 7

0 ± 3 6 6 6 6 6 6 6 6

4 7 7 7 7 7 7 7 7

5 ± 31 9 9 9 9 9 9 9 9





Modulation selection Modulation selection is based on 8-PSK MEAN_BEP, 8-PSK

CV_BEP and GMSK_MEAN BEP according to the table below.

If MS does not support 8-PSK in the uplink, GMSK shall bechosen in the uplink algorithm.

The items in the table above are the 8PSK MEAN_BEP and CV_BEPvalues. The table is used as follows: the algorithm locates an entry inthe table based on measured 8PSK MEAN_BEP and CV_BEP values.This entry is compared to the measured GMSK MEAN_BEP value. If the measured value is larger GMSK is chosen, otherwise 8PSK ischosen.

8PSK CV_BEP-class8PSKMEAN_BEP-class

1 2 3 4 5 6 7

0 4 4 4 4 4 4 4

1 6 6 6 6 6 5 5

2 9 9 9 9 9 7 6

3 ± ± ± 21 12 11 8

4 ± ± ± ± 20 13 12

5 ± ± ± ± 24 21 21

6 ± 31 ± ± ± ± ± ± ±





The figure shows theperformance of a BEP basedLA algorithm.

Throughput vs. C/I in TU3ideal FH.

Realistic IR memory hasbeen used.

The algorithm works by usingBEP reports from MS toselect the coding scheme.

0

10

20

30

40

50

60

0 5 10 15 20 25 30

MCS-1MCS-2

MCS-3

MCS-4

MCS-5

MCS-6

MCS-7

MCS-8MCS-9

LA

Nokia EGPRS3 Details

Documents

Transcript of Nokia EGPRS3 Details