umt irc app 016664 - hsdpa engineering guide v01.06 ext nortel

Upload: cmoorescribd

Post on 14-Apr-2018

216 views

Category:

Documents


0 download

TRANSCRIPT

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    1/141

    HSDPA Engineering Guide

    Document number: UMT/IRC/APP/016664Document issue: 01.06 / ENDocument status: Preliminary

    Date: 20/Jan/2006

    External Document

    Copyright 2005 Nortel Networks, All Rights Reserved.

    Printed in France

    NORTEL CONFIDENTIAL

    The information contained in this document is the property of Nortel Networks. Except as specifically authorized in

    writing by Nortel Networks, the holder of this document shall keep the information contained herein confidential

    and shall protect same in whole or in part from disclosure and dissemination to third parties and use same forevaluation, operation and maintenance purposes only.

    The content of this document is provided for information purposes only and is subject to modification. It does not

    constitute any representation or warranty from Nortel Networks as to the content or accuracy of the information

    contained herein, including but not limited to the suitability and performances of the product or its intended

    application.

    This is the Way. This is Nortel, Nortel, the Nortel logo, and the Globemark are trademarks of Nortel Networks. All

    other trademarks are the property of their owners.

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    2/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 2/141

    PUBLICATION HISTORY

    20/Jan/2005

    Issue 01.06 / EN, Preliminary

    Document update / UA4.2 CuR

    25/Nov/2005

    Issue 01.05 / EN, Draft

    Document update / External Version

    18/Oct/2005

    Issue 01.04 / EN, Draft

    Document updated after review for external version / External Version

    04/Oct/2005

    Issue 01.03 / EN, Draft

    Document updated for W.39 Load / External Version

    02/Sep/2005

    Issue 01.02 / EN, Draft

    Document updated after review

    19/Aug/2005

    Issue 01.01 / EN, Draft

    Document Creation

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    3/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 3/141

    CONTENTS

    1 INTRODUCTION............................................................................................................................88H

    91.1 OBJECT....................................................................................................................................89H9

    1.2 SCOPE OF THIS DOCUMENT .......................................................................................................90H9

    1.3 NOMENCLATURE.......................................................................................................................91H9

    2 RELATED DOCUMENTS............................................................................................................92H11

    2.1 REFERENCE DOCUMENTS ........................................................................................................93H11

    3 HSDPA OVERVIEW ....................................................................................................................94H12

    3.1 SYSTEM OVERVIEW .................................................................................................................95H

    123.1.1 New transport and physical channels ...........................................................................96H143.1.2 Fast link adaptation.......................................................................................................97H163.1.3 Fast Retransmission Mechanism (HARQ)....................................................................98H173.1.4 Fast scheduling.............................................................................................................99H22

    3.2 DEPLOYEMENT SCENARIOS .....................................................................................................100H26

    3.2.1 Dual Carrier ...................................................................................................................101H263.2.2 Single carrier.................................................................................................................102H27

    3.3 HSDPARESOURCES..............................................................................................................103H27

    3.3.1 OVSF Codes.................................................................................................................104H273.3.2 Power ............................................................................................................................105H283.3.3 HSDPA Channels & CQI...............................................................................................106H29

    3.4 UECATEGORIES ....................................................................................................................107H36

    3.5 CALL MANAGEMENT................................................................................................................108H37

    3.5.1 Traffic segmentation......................................................................................................109H383.5.2 HSDPA CAC .................................................................................................................110H413.5.3 Call Setup (Dataflow) ....................................................................................................111H433.5.4 Call Release (Dataflow) ................................................................................................112H453.5.5 HSDPA related Transitions ...........................................................................................113H46

    4 HSDPA CONFIGURATION .........................................................................................................114H51

    4.1 RANMODEL AND PARAMETERS ..............................................................................................115H51

    4.1.1 RRM Impact ..................................................................................................................116H

    514.1.2 BTS Impact....................................................................................................................117H644.1.3 IuB Impact.....................................................................................................................118H65

    4.2 HSDPAACTIVATION...............................................................................................................119H66

    6 UA4.2 HSDPA SPECIFIC FEATURES & IMPACT ON EXISTING ALGORITHMS...................120H68

    6.1 ALWAYS ON ...........................................................................................................................121H68

    6.1.1 Mechanism....................................................................................................................122H686.1.2 Activation & Mode .........................................................................................................123H686.1.3 Reminder for timer usage..............................................................................................124H726.1.4 Parameters Settings and Recommendations ...............................................................125H72

    6.2 IRMALGORITHMS...................................................................................................................126H73

    6.2.1 Irm Scheduling Downgrade/Upgrade Interworking .......................................................127H736.2.2 iRM Cac/iRM Pre-Emption Interworking .......................................................................128H73

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    4/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 4/141

    6.2.3 RB Adaptation Interworking ..........................................................................................129H736.3 MOBILITY PROCEDURES ..........................................................................................................130H73

    6.3.1 Intra-Frequency Mobility................................................................................................131H746.3.2 Inter-Frequency Mobility................................................................................................132H766.3.3

    Inter-System Mobility.....................................................................................................133H76

    6.3.4 U-Plane Traffic Handling...............................................................................................134H776.3.5 Summary of inter-frequency and inter-RAT scenarios..................................................135H78

    6.4 POWER MANAGEMENT ............................................................................................................136H78

    6.4.1 Introduction....................................................................................................................137H786.4.2 Flexible power management feature.............................................................................138H786.4.3 HS-SCCH power control feature...................................................................................139H946.4.4 Parameters Settings and Recommendations ...............................................................140H95

    6.5 TRANSPORT ...........................................................................................................................141H98

    6.5.1 Iub Bandwidth Limitation: Why this feature is needed?................................................142H986.5.2 Feature Description.......................................................................................................143H996.5.3 Case of Drift Iur .......................................................................................................... 144H101

    6.5.4 Parameters Settings and Recommendations ............................................................145H

    102

    7 HSDPA THROUGHPUT OPTIMIZATION ................................................................................ 146H108

    7.1 POWERALLOCATION AND USER THROUGHPUT...................................................................... 147H108

    7.2 CQISELECTION AT UE ........................................................................................................ 148H111

    7.3 CQI ADJUSTMENT BASED ON BLER...................................................................................... 149H111

    7.4 BTS CREDIT ALLOCATION ..................................................................................................... 150H113

    7.5 HARQ RETRANSMISSIONS ................................................................................................... 151H113

    7.6 RLCSETTINGS FOR ULDCH............................................................................................... 152H114

    7.7 DLRLC SETTINGS FOR HSDPA........................................................................................... 153H117

    7.8 OPTIMISED TCP SETTINGS ................................................................................................... 154H117

    7.9 PARAMETERS SETTINGS ...................................................................................................... 155H118

    8 HSDPA CAPACITY ASPECTS ................................................................................................ 156H119

    8.1 CEMCAPACITY ................................................................................................................... 157H119

    8.1.1 Product limits.............................................................................................................. 158H1198.1.2 Capacity analysis ....................................................................................................... 159H119

    9 PRODUCT RECOMMENDATIONS.......................................................................................... 160H129

    9.1 HSDPA COMPATIBILITY WITH UTRANNETWORK ELEMENTS .................................................161H

    1299.1.1 RNC............................................................................................................................ 162H1299.1.2 BTS ............................................................................................................................ 163H129

    9.2 HSDPA COMPATIBILITY WITH MODULES ............................................................................... 164H130

    9.2.1 RNC............................................................................................................................ 165H1309.2.2 BTS ............................................................................................................................ 166H130

    9.3 HSDPASYSTEMIMPACT................................................................................................ 167H131

    9.3.1 RNC functions............................................................................................................ 168H1319.3.2 BTS functions............................................................................................................. 169H131

    9.4 HSDPA AND UTRANINTERFACES ...................................................................................... 170H135

    9.4.1 Radio Interface........................................................................................................... 171H135

    9.4.2 IuB..............................................................................................................................172H

    1359.4.3 Iu-CS .......................................................................................................................... 173H1369.4.4 Iu-PS .......................................................................................................................... 174H136

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    5/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 5/141

    9.4.5 Iu-R............................................................................................................................. 175H1369.4.6 Iu-PC .......................................................................................................................... 176H136

    10 ABBREVIATIONS AND DEFINITIONS.................................................................................... 177H137

    10.1 ABBREVIATIONS

    ...................................................................................................................178H

    13710.2 DEFINITIONS........................................................................................................................ 179H140

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    6/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 6/141

    TABLES

    0HTable 1: Number of processes per UE category 180H171H

    Table 2: RV coding for 16QAM181H

    182HTable 3: RV coding for QPSK 182H183HTable 4: RV update table in the MIR case (Trv[i]) 183H194HTable 5: RV update table in the PIR case (Trv[i]) 184H195HTable 6: UE capabilities 185H376HTable 7: RB Configuration and system behaviour 186H387HTable 8: RRC Connection Request and suitable layer 187H408HTable 9: HSDPA related Transitions 188H479HTable 10 : Always-on timer usage 189H7210HTable 11: CQI update summary 190H9111HTable 12: CQI Mapping 191H9212HTable 13: HS-SCCH power offset table according the averaged CQI 192H9513HTable 14: UL rate required vs. DL rate 193H11714HTable 15: Parameters summary for optimized throughput 194H11815HTable 16: H-BBU limitations 195H119

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    7/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 7/141

    FIGURES

    16HFigure 1: R99 principle .......................................................... ................................................................... ............ 196H1217H

    Figure 2: HSDPA principle...................................................................................................................................197H

    1218HFigure 3: HSDPA layer2/layer1 flows ............................................................ ...................................................... 198H1319HFigure 4: Mac-hs entity on UTRAN side.............................................................................................................. 199H1320HFigure 5: Transport channel configuration............................................................................................................ 200H1521HFigure 6: HSDPA channels and associated R99 channels..................................................................................... 201H1522HFigure 7: Example of AMC : Throughput versus Ior/Ioc (radio condition) .......................................................... 202H1623HFigure 8: RV parameters assignment algorithm.................................................................................................... 203H2024HFigure 9: ACK/NACK/DTX management for HARQ processes.......................................................... ................ 204H2125HFigure 10: Scheduler overview ................................................................. ............................................................ 205H2326HFigure 11: HSDPA on dedicated layer.................................................................................................................. 206H2627HFigure 12: Example of OVSF tree usage with HSDPA ............................................................................ ............ 207H2728HFigure 13: OVSF allocation strategy..................................................................................................................... 208H2829HFigure 14: Timing relationship at NodeB between physical channels ................................................................. . 209H2930HFigure 15: HS-SCCH structure ............................................................................ ................................................. 210H3031HFigure 16: HS-PDSCH structure........................................................................................................................... 211H3032HFigure 17: HS-DPCCH structure ................................................................... ....................................................... 212H3133HFigure 18: CQI Processing.................................................................................................................................... 213H3234HFigure 19: CQI offset computation based on BLER ..................................................................... ........................ 214H3435HFigure 20: Scheduler and Flow Control disabled.................................................................................................. 215H3636HFigure 21: HSDPA Call setup / initial connection (Cell_DCH)............................................................................ 216H4337HFigure 22: HSDPA Call setup / RAB allocation phase (call establishment done on DCH).................................. 217H4438HFigure 23: Call release (RAB release case)........................................................................................................... 218H4639HFigure 24: RRM new HSDPA subtree ................................................................ .................................................. 219H5140HFigure 25: FDDCell HSDPA Related new objects ................................................................... ............................ 220H5141HFigure 26: HSDPARncConf subtree ....................................................................... .............................................. 221H5342HFigure 27: HSDPA Radio Bearer subtree .................................................................. ........................................... 222H5543HFigure 28: HSDSCHx4SRBDCCH3_4K subtree......................................................................................... ......... 223H5844HFigure 29: HSDPA RLC subtree........................................................................................................................... 224H6045HFigure 30: HSDPA UsHoConf subtree ..................................................................... ............................................ 225H6146HFigure 31: HSDPA BTS new object ...................................................................... ............................................... 226H6447HFigure 32: IuB HSDPA new object....................................................................................................................... 227H6548HFigure 33: Always-on for HSDPA (degraded2AlwaysOn mode) .................................................................... ..... 228H7049HFigure 34: Always-on for HSDPA (releaseOnly mode)........................................................................................ 229H7150HFigure 35: HS-DSCH link is deleted and re-established on new primary (nominal case) .................................... 230H7551HFigure 36: Summary of Inter-freq/inter-RAT mobility......................................................................................... 231H7852HFigure 37: Power allocation at RNC level ................................................................... ......................................... 232H7953HFigure 38: Physical shared channel reconfiguration message............................................................................... 233H8154HFigure 39: Power allocation at NodeB level ................................................................. ........................................ 234H8155H

    Figure 40: Transmitted carrier power (on the left) and averaged HSDPA power (on the right)...........................235H

    8356HFigure 41: Common measurement .............................................................. .......................................................... 236H8457HFigure 42: Power measurement process................................................................................................................ 237H8558HFigure 43: Power distribution between HS-DSCH and HS-SCCH channels........................................................ 238H8659HFigure 44: measurementPowerOffset transmission......................... ..................................................................... . 239H8760HFigure 45: HS-DSCH power management for 1st transmission............................................................................ 240H8961HFigure 46: Mac-hs transport block adaptation according to the number of Mac-d PDU to transmit .................... 241H9162HFigure 47: HS-DSCH power management for retransmission................................................................. ............. 242H9363HFigure 48: Remaining power for multi-users per TTI scheduling......................................................... ................ 243H9464HFigure 49: HS-SCCH power control overview .................................................................... ................................. 244H9565HFigure 50: discard and backpressure thresholds.................................................................................................... 245H9966HFigure 51: example of traffic regulation with backpressure................................................................................ 246H10167HFigure 52: feature behaviour on Iur ........................................................................ ............................................ 247H10268HFigure 53: example of transport topologies "with ATM priority"....................................................................... 248H10269HFigure 54: example of transport topology "without ATM priority".................................................................... 249H10370HFigure 55: User multiplexing in Code and Time domains .................................................................... .............. 250H108

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    8/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 8/141

    71HFigure 56: User throughput vs. HSDPA power, UE cat 6................................................. .................................. 251H10972HFigure 57: HARQ BLER vs. User RLC Throughput, UE cat 12, drive test........ ................................................ 252H11173HFigure 58: User throughput vs. BLER on HS-PDSCH, UE cat 12 ................................................................. .... 253H11274HFigure 59: User throughput vs. status prohibit timer, UE cat 6........................................................................... 254H11575HFigure 60: User throughput vs. polling timer vs. UL BLER............................................................................... 255H11576HFigure 61: User throughput vs. receive window size at UE........ ............................................................... ......... 256H11677HFigure 62: H-BBU resource allocation modes.................................................................. .................................. 257H12078HFigure 63: iCEM consumption for a PS RB DL 128 kbps / UL 384 kbps (R99 Case) ....................................... 258H12179HFigure 64: iCEM consumption for a PS RB HSDPA / UL 128 kbps (HSDPA Case)......................................... 259H12280HFigure 65 : Comparison between the CEM R99 Capacity and the CEM HSDPA Capacity for same input traffic

    and same CEM configuration (same hardware) ................................................................. ......................... 260H12281HFigure 66: CEM Capacity vs. HSDPA Penetration.................................................. ........................................... 261H12382HFigure 67: CEM Admission Blocking type (R99 or HSDPA) versus HSDPA Penetration Factor..................... 262H12483HFigure 68: CEM Capacity figures for different configurations depending on the HSDPA penetration factor (UL

    64 kbps)....................................................................................................................................................... 263H12484HFigure 69: Admission blocking by service vs HSDPA penetration factor.......................................................... 264H12685HFigure 70: CEM Capacity figures for different configurations depending on the HSDPA penetration factor (UL

    128 kbps)..................................................................................................................................................... 265H127

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    9/141

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    10/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 10/141

    Parameter Object

    Range & Unit

    User

    Class

    Granularity

    Value

    The protocol messages are written in CAPITAL LETTERS.

    The Information Elements (IE) contained in the protocol messages are written the

    following way: TPC_DL_Step_Size .

    The datafill rules are presented as the following:

    Rule:

    The system restrictions are presented as the following:

    Restriction:

    The engineering recommendations on parameter value are presented as the

    following:

    Engineering Recommendation:

    Some parameters values can not be provided in this document; in that case, the

    following abbreviations are used:

    o N.A.: Not Applicable.

    o N.S.: Not Significant.

    o O.D.: Operator Dependant (depends on operator network specific

    configuration. Example: addressing parameters).

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    11/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 11/141

    2 RELATED DOCUMENTS

    2.1 REFERENCE DOCUMENTS

    [R1] UMT/DCL/DD/0020 UTRAN Parameters User Guide

    [R2] 3GPP TS 25.308 UTRA High Speed Downlink Packet Access

    (HSPDA); Overall description; Stage 2

    [R3] 3GPP TS 34.108 Common Test Environments for User Equipment

    (UE) Conformance Testing

    [R4] 3GPP TS 25.212 Multiplexing and channel coding (FDD)

    [R5] 3GPP TS 25.214 Physical layer procedures (FDD)

    [R6] UMT/BTS/DD/012447 NodeB HSDPA functional specification

    Commercial release

    [R7] UMT/BTS/DD/008196 Mac-hs Scheduler

    [R8] UMT/BTS/DD/012545 Variable Length CQI averaging

    [R9] UMT/IRC/APP/0164 Iub transport Engineering Guide

    [R10] NTP 411-8111-575 HSDPA Feature Activation Procedure

    [R11] UMT/SYS/DD/013319 HSDPA System Specification

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    12/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 12/141

    3 HSDPA OVERVIEW

    3.1 SYSTEM OVERVIEW

    3GPP has standardized HSDPA in Release 5 [ 279HR2] in order to increase maximum user

    throughput for downlink packet data (streaming, interactive and background traffic

    classes) and decrease downlink packet transmission delay. This Release 5 is fully

    compatible with the previous Release 99 (R99).

    In R99, data are transmitted on a dedicated channel with a given user throughput and

    a downlink transmitted power controlled according to the radio conditions:

    PowerPower

    ControlControl

    Data Power

    Unused Power Data

    Unused

    Same Throughput

    Figure 1: R99 principle

    In HSDPA, data are transmitted on a shared channel by using all the available power

    and by controlling the downlink user throughput according to the radio conditions:

    RateRate

    AdaptationAdaptation 100% Power

    100%

    Figure 2: HSDPA principle

    Typically, a user in good radio conditions will receive his data with a high bit rate

    whereas a user in bad radio conditions will receive his data with a lower bit rate.

    The efficiency of this rate adaptation is due to a new MAC entity, the Mac-hs layer,

    located in the NodeB (see the two following figures), near the physical channel, which

    allows a high reactivity in the resource allocation according to the RF conditions

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    13/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 13/141

    changes. This Mac-hs layer manages the scheduling of users and the retransmissions

    of packets.

    HS-DSCH

    AssociatedUplink

    Signaling

    AssociatedDownlink

    Signaling

    DCCH DTCHDTCHMAC Control MAC ControlCCCH CTCHBCCHPCCHMAC Control

    RRC (RNC)RRC (RNC)

    RLC (RNC)RLC (RNC)

    HS-PDSCH

    FACH

    S-CCPCH

    FACH

    S-CCPCH

    RACH

    PRACH

    RACH

    PRACH

    DSCH

    PDSCH

    DSCH

    PDSCH

    DCH

    DPCH

    CPCH

    PCPCH

    CPCH

    PCPCH

    PCH

    S-CCPCH

    PCHPCH

    S-CCPCHHS-DPCCHHS-SCCH

    MAC-c/sh

    (C-RNC)

    MAC-c/sh

    (C-RNC)

    DCH

    DPDCH/DPCCH

    R99 L1: Channel Coding / Multiplexing (NodeB)R99 L1: Channel Coding / Multiplexing (NodeB)R5 L1: HSDPA (NodeB)R5 L1: HSDPA (NodeB)

    MAC-d

    (S-RNC)MAC-hs

    (NodeB)

    MAC-hs

    (NodeB)

    Figure 3: HSDPA layer2/layer1 flows

    Figure 4: Mac-hs entity on UTRAN side

    HSDPA benefits are based on:

    New transport and physical channels.

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    14/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 14/141

    Fast link adaptation.

    Fast retransmission mechanism (HARQ).

    Fast scheduling.

    3.1.1 NEW TRANSPORT AND PHYSICAL CHANNELS

    In R99, downlink data are sent on a DCH (Dedicated CHannel) which is mapped on

    the DPDCH (Dedicated Physical Data CHannel). In HSDPA, downlink data are sent

    on a HS-DSCH (High Speed Downlink Shared CHannel) which is mapped on one or

    several HS-PDSCH (High Speed Physical Downlink Shared CHannel). Users are

    multiplexed on the HS-DSCH channel in time and code. Transmission is based on

    shorter sub-frames of 2ms (TTI) instead of 10ms in R99.

    In downlink, the HS-PDSCH are transmitted with the HS-SCCH (High Speed Shared

    Control CHannel) channel. This channel is broadcasted over the cell but his

    information concerned only the user who has to receive the HS-PDSCH. The HS-

    SCCH allows the user to know if the HS-PDSCH are for him and to decode them

    correctly.

    Radio conditions information and acknowledgement are reported by the UE to the

    NodeB through the HS-DPCCH channel. This channel allows the NodeB to adapt the

    downlink data rate and to manage retransmission process. The HS-DPCCH is dividedin two parts. The first one is the Channel Quality Indicator (CQI) which is a value

    between 1 and 30 characterizing the radio conditions (1 = bad radio conditions and 30

    = good radio conditions). The second one is the acknowledgement information: if data

    are well received by the UE, the UE sends to the NodeB an Ack, otherwise a Nack.

    HS-DSCH channel is always associated to a DCH. This induces the following

    transport channel configuration for any UE established in HSDPA (see the following

    figure):

    one DCH handling the signaling in both UL and DL,

    one DCH transporting the UL traffic,

    one HS-DSCH for the DL traffic.

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    15/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 15/141

    Figure 5: Transport channel configuration

    The following figure summarizes the different channels needed for a HSDPA call:

    NodeB

    HSDPA UE

    HS-PDSCH for data (I/B) trafficHS-PDSCH for data (I/B) traffic

    HSDPA channelsHSDPA channels

    HS-SCCH signaling part (UE id, ) associated

    to HS-PDSCH

    HS-SCCH signaling part (UE id, ) associated

    to HS-PDSCH

    HS-DPCCH Feedback informationHS-DPCCH Feedback information

    Associated DPCH for data, speech + SRB trafficAssociated DPCH for data, speech + SRB traffic

    Figure 6: HSDPA channels and associated R99 channels

    In this release of HSDPA, UE categories 6 and 12 are supported, allowing a maximum

    data rate in downlink of respectively 3.65Mbit/s and 1.8Mbit/s. As a consequence the

    following radio bearers are be supported:

    Interactive or background / UL:8 DL: [max bit rate for UE categories 12 and 6]

    / PS RAB + UL:3.4 DL:3.4 kbps SRBs for DCCH (see below)

    Interactive or background / UL:32 DL: [max bit rate for UE categories 12 and

    6] / PS RAB + UL:3.4 DL:3.4 kbps SRBs for DCCH

    Interactive or background / UL:64 DL: [max bit rate for UE categories 12 and

    6] / PS RAB + UL:3.4 DL:3.4 kbps SRBs for DCCH

    Interactive or background / UL:128 DL: [max bit rate for UE categories 12 and

    6] / PS RAB + UL:3.4 DL:3.4 kbps SRBs for DCCH

    Interactive or background / UL:384 DL: [max bit rate for UE categories 12 and6] / PS RAB + UL:3.4 DL:3.4 kbps SRBs for DCCH

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    16/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 16/141

    The maximum bit rate that can be achieved in the UL can be the bottleneck for the

    maximum bit rate achievable in the DL. For instance, excessive delay of RLC/TCP

    acknowledgements due to low bandwidth in the UL will limit the DL throughput, even if

    the RF conditions would allow more.

    In UA04.2, Nortel implements the RB adaptation feature that dynamically adapts the

    UL bit rate to the amount of traffic carried over the RB. UL adaptation ranges from

    32kbps up to 384kbps, but 8kbps is not eligible. Therefore, although UL:8 DL:[max bit

    rate for UE categories 12 and 6] will be allocated by the RNC if UL:8 is explicitly

    requested in the RAB assignment, it is not recommended to do so, otherwise the user

    will experience low throughput in the DL.

    3.1.2 FAST LINK ADAPTATION

    Every TTI, Adaptive Modulation and Coding (AMC) is updated according to the radio

    conditions experienced by the UE and his category (see UE categories section).

    AMC (number of codes, code rate and modulation type) is chosen among 30

    possibilities corresponding to one CQI in order to reach the maximum bit rate while

    guarantying a certain QoS (10% BLER for example). All UE categories have to

    support QPSK and 16QAM modulation except categories 11 and 12 which only

    support QPSK (16QAM modulation allowing higher bit rate). The following figure

    illustrates the AMC by showing the throughput versus the radio conditions (Ior/Ioc):

    QPSK

    QPSK

    QPSK

    16QAM

    16QAM

    -20 -15 -10 -5 0 50

    100

    200

    300

    400

    500

    600

    700

    800

    Ior/Ioc (dB)

    Throughput(kbps)

    AMC Illustration

    QPSK

    QPSK

    QPSK

    16QAM

    16QAM

    QPSK

    QPSK

    QPSK

    16QAM

    16QAM

    -20 -15 -10 -5 0 50

    100

    200

    300

    400

    500

    600

    700

    800

    Ior/Ioc (dB)

    Throughput(kbps)

    AMC Illustration

    Figure 7: Example of AMC : Throughput versus Ior/Ioc (radio condition)

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    17/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 17/141

    3.1.3 FAST RETRANSMISSION MECHANISM (HARQ)

    The HARQ (Hybrid Automatic Repeat Query) is a retransmission mechanism which

    consists in:

    retransmitting by the NodeB the data blocks not received or received with

    errors by the UE;

    combining by the UE the transmission and the retransmissions in order to

    increase the probability to decode correctly the information.

    3.1.3.1 NUMBER OF HARQ PROCESSES

    There is an HARQ process assigned per transport block for all the transmissions. The

    number of processes per UE is limited and depends on its category. The number of

    processes per UE category is the one given in [ 280HR3]:

    Ue Category 1 2 3 4 5 6 7 8 9 10 11 12

    Number of HARQ Processes 2 2 3 3 6 6 6 6 6 6 3 6

    Table 1: Number of processes per UE category

    Once this number is reached, the UE should not be eligible by the scheduler for new

    transmissions unless one of them is reset (ACK reception, discard timer expiration,max number of retransmissions reached).

    3.1.3.2 RV PARAMETERS

    The IR (Incremental Redundancy) and modulation parameters necessary for the

    channel coding and modulation steps are: the r, s and b values. The r and s

    parameters (Redundancy Version or RV parameters) are used in the second rate

    matching stage, while the b parameter is used in the constellation rearrangement step

    (see [281HR4] for details):

    s is used to indicate whether the systematic bits (s=1) or the non-systematic

    bits (s=0) are prioritized in transmissions.

    r(range 0 to rmax-1) changes the initialization Rate Matching parameter value

    in order to modify the puncturing or repetition pattern.

    The b parameter can take 4 values (0, , 3) and determines which operations

    are produced on the 4 bits of each symbol in 16QAM. This parameter is not

    used in QPSK and constitutes the 16QAM constellation rotation for averaging

    LLR at the turbo decoder input.

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    18/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 18/141

    These three parameters are indicated to the UE by the Xrv value sent on the HS-

    SCCH (see section 282H3.3.3 "283HHSDPA Channels & CQI). The coding tables of Xrv are

    given hereafter:

    Xrv (Value) s r b

    0 1 0 0

    1 0 0 0

    2 1 1 1

    3 0 1 1

    4 1 0 1

    5 1 0 2

    6 1 0 3

    7 1 1 0

    Table 2: RV coding for 16QAM

    Xrv (Value) s r

    0 1 0

    1 0 0

    2 1 1

    3 0 1

    4 1 2

    5 0 2

    6 1 3

    7 0 3

    Table 3: RV coding for QPSK

    The determination of the s, r and b parameters will be based on the Xrv update, but

    not necessarily in the increasing order. The update indeed follows a predefined order

    stored in a table (called hereafter Trv). The only requirement to fill this table is that

    Trv[0] = 0 for QPSK, or Trv[0] = 0, 4, 5 or 6 for 16QAM (s = 1 and r = 0 must be the

    nominal configuration).

    A configurable parameter (CC/PIR/MIR) indicates the possibility of switching between

    Chase Combining, a Partial IR or a mix between Partial and Full IR sequence. Itimplies that 3 different tables must be stored (see below), chosen accordingly:

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    19/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 19/141

    The Chase Combining option corresponds to the first redundancy version

    always applied for all (re)transmissions.

    The PIR indicates that for all redundancy versions, the systematic bits must

    be transmitted (blocks are self-decodable). Only the RV with s = 1 must be

    taken into account.

    The MIR corresponds to a sequence where both systematic and

    nonsystematic bits can be punctured. All possible redundancy versions are

    assumed and it corresponds to default Nortels version.

    i 1 2 3 4 5 6 7 8

    Xrv(QPSK) 0 2 5 6 1 3 4 7

    Xrv(16QAM) 6 2 1 5 0 3 4 7

    Table 4: RV update table in the MIR case (Trv[i])

    i 1 2 3 4 5 6

    Xrv(QPSK) 0 2 4 6

    Xrv(16QAM) 6 2 5 0 4 7

    Table 5: RV update table in the PIR case (Trv[i])

    The rules to compute the Xrv parameters then are (see the following figure):

    For the first transmission, Xrv is initialized to Trv[0].

    Upon reception of a NACK, Xrv is assigned the next value in the table (once

    the last value of the table, Nmax, has been set, the next value should be the

    first one again).

    In case of no reception of ACK/NACK (DTX indication), the parameters must

    not be updated so that the same information not received by the UE should be

    sent again (this ensure no systematic bits are lost, because all blocks may not

    be self-decodable).

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    20/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 20/141

    New transmission ?Xrv = Trv[0]

    k = 0

    Y

    N

    DTX indication ? Xrv(n) = Xrv(n-1)Y

    N

    k = k + 1

    Xrv(n) = Trv[k mod Nmax]Nmax = 1 (CC)

    = 4 (PIR - QPSK)

    = 6 (PIR 16QAM)

    = 8 (MIR)

    Figure 8: RV parameters assignment algorithm

    3.1.3.3 STATE OF HARQ PROCESSES

    The following figure describes the different states of HARQ processes and possible

    actions related to these.

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    21/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 21/141

    ACK/NACK/DTX ?

    HARQ process assigned

    by the scheduler

    Y

    Update of RV parameters

    Data transmission

    Wait for ACK/NACK

    reception

    Insertion of DTX

    indication

    Reset HARQ processRemove Mac-d PDU

    Update structures

    Nret = Nret +1

    Nret > Nret_max ?

    Wait for

    retransmission

    NACK

    DTX

    N

    WACK state

    NACK/DTX state

    ACK

    Figure 9: ACK/NACK/DTX management for HARQ processes

    Once a UE is scheduled, an HARQ process is assigned that may correspond to either

    a new Transport Block or a retransmission. The RV parameters are computed

    accordingly as described before (see RV PARAMETERS section) and data is

    transmitted. The HARQ process is then waiting for feedback information

    (ACK/NACK/DTX) and is set in the so-called WACK state (Waiting for

    Ack/Nack/DTX). The exact timing for reception of the feedback information must be

    computed thanks to the chip offset and relatively to the TTI corresponding to the

    transmission.

    Upon reception of the feedback information, three behaviors occur:

    In case of an ACK, the HARQ process is reset and corresponding Mac-d

    PDUs are removed from memory. This HARQ process can now be used for a

    new transmission.

    In case of a NACK, the number of retransmissions must be incremented. If the

    maximum number of retransmissions is not reached, the HARQ process is set

    in the so-called NACK state and then inserted in the NACK list of HARQ

    processes.

    In case of a DTX indication, the same actions as for a NACK reception aredone, except that a parameter must be updated to notify DTX detection (this

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    22/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 22/141

    changes the RV parameter update, see RV PARAMETERS section). The

    process is then set in the DTX state.

    The processes in the NACK or DTX state are just waiting for being re-scheduled for anew retransmission.

    3.1.4 FAST SCHEDULING

    3.1.4.1 PRINCIPLES

    The aim of the Mac-hs scheduler is to optimize the radio resources occupancy

    between users. Every TTI, it must then select Queue IDs for which data is going to be

    transmitted and the amount of corresponding Mac-d PDUs to transmit.

    The scheduler first receives as input every TTI the number of codes available and the

    remaining power for HS-PDSCH and HS-SCCH (see POWER MANAGEMENT

    section). The received ACK/NACK and CQI must also be provided to the scheduler

    when available. Thanks to this information, UE capabilities, configuration parameters

    provided by the RNC and taking into account the previously scheduled data, the

    scheduler can select the subflows of the users to schedule in order to optimally use

    available resources. The main concepts of the scheduler are:

    Retransmissions are of higher priority than new transmissions and should be

    scheduled first.

    Users with higher priority (CmCH-PI) and better CQI are favoured and get

    most part of the bandwidth.

    The transport blocks should always be optimized according to the transmitted

    CQI when possible (if enough codes and power are available and if theres no

    CPU limitation).

    No queue ID should be left starving, i.e. the scheduler should always allocate

    even a small part of radio resources to all users (even those with low priority

    and bad CQI).

    3.1.4.2 ALGORITHM

    The architecture of the scheduler is presented hereafter (see [284HR7] for more details on

    the algorithm):

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    23/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 23/141

    Figure 10: Scheduler overview

    Before entering the scheduler, the Mac-d PDUs for each queue ID of each user are

    classified according to their corresponding priority (CmCH-PI). Every TTI, the

    scheduler is launched in order to efficiently assign the available resources (number of

    codes and remaining power) to the different users.

    The first step consists in managing the retransmissions. The NACKed blocks are

    scheduled first, in a FIFO order when possible (in case the UE capabilities prevent

    from receiving data in the corresponding TTI, it is not retransmitted in that TTI). Then,

    once retransmissions are handled, the remaining number of codes and power are

    computed and constitute the input of the next step.

    The scheduling of first-transmitted data is based on a two-stage solution:

    The first stage selects one priority queue among the active ones.

    The second stage consists in selecting the user(s) within the chosen priority

    queue to schedule.

    These two steps are repeated as long as some resources remain (codes, power and

    CPU) and if data can be scheduled for the corresponding TTI.

    3.1.4.3 FIRST STAGE

    The following paragraph describes the general behavior. In this version, as only onepriority queue is available, this first stage is transparent.

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    24/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 24/141

    The selection of a priority queue is based on a cost function that takes into account

    credits assigned to each priority queue and the number of Mac-d PDUs already

    transmitted in the last TTI for these queues. The aim is to provide some throughput

    per queue according to its priority: the higher the priority, the higher the allocated

    bandwidth.

    The credits per queue then depend on their respective priority, on the number of

    active priority queues but shall also be proportional to the number of active QIDs per

    priority queue (in order to ensure the bandwidth of high priority QIDs remains more

    important that low priority ones independently on the number of UEs per priority

    queue).

    The credits per priority queues are recomputed any time the status of a priority queue

    changes (active/inactive) or if the number of active QIDs in a queue changes. When

    credits are recomputed, the ratio between queues initially setup is kept.

    3.1.4.4 SECOND STAGE

    Once a priority queue has been selected, one or more users within that queue are

    scheduled. The selection of one QID is done according to another cost function that

    takes into account the processed CQI (noted CQIprocessed in CQI section) and the

    number of Mac-d PDUs transmitted in the last TTI. This ensures that all users areselected but that the bandwidth allocated to the priority queue is separated between

    users according to their CQI (the higher the CQI, the higher the available throughput).

    A user can only be considered as candidate if it is allowed to receive some data in the

    current TTI, i.e. if several criteria are respected (CQIprocessed 0, min inter TTI distance,

    AckNack repetition factor, one HARQ process available, UE not already scheduled for

    retransmissions).

    For each candidate user, HS-SCCH power is determined (see POWER

    MANAGEMENT section) and power checking is done on both the current TTI and the

    previous one (HS-SCCH and corresponding HS-PDSCH only overlap by 1 slot). If

    there is not enough power to transmit the HS-SCCH in the current TTI, the user is not

    selected.

    Then for the UE with the lowest cost within the remaining candidates UEs, the number

    of PDU to transmit as well as the number of codes and the transmitted power are

    chosen according to the processed CQI in order to fit as well as possible to what the

    UE can correctly receive (see POWER MANAGEMENT section for more details on

    the algorithm):

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    25/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 25/141

    If there doesnt remain enough power to transmit the HS-PDSCH, another

    configuration requiring less power is selected if possible (transport block size

    reduced, corresponding to a smaller CQI referred as CQIapplied in Section 285H6.4

    286HPower Management.

    If there doesnt remain enough codes to transmit the HS-PDSCH, the

    configuration is changed (transport block size reduced, corresponding to a

    smaller CQI) to use the remaining codes. The power is then updated

    accordingly.

    Anytime a UE is scheduled, its cost is recomputed according to the transmitted

    number of MAC-d PDUs.

    Another priority queue must be selected when no more QID in the current priority

    queue can be scheduled (all HARQ processes used for the corresponding UE, min

    inter TTI distance not compliant with that TTI, no other QID). As only one priority

    queue is available in this version, the first stage is not recalled and the TTI processing

    is considered as complete.

    The cost of each QID of the selected priority queue is updated anytime another queue

    must be selected (according to previous criteria) or at the end of a TTI (no more HS-

    SCCH/HS-PDSCH code, power or CPU).

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    26/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 26/141

    3.2 DEPLOYEMENT SCENARIOS

    3.2.1 DUAL CARRIER

    The preferred scenario is to deploy a new layer dedicated to HSDPA on a new

    frequency. This layer may be deployed either widely or restricted to some hot-spots.

    Layer with HSDPA configured

    Layer without HSDPA

    Figure 11: HSDPA on dedicated layer

    The advantage of this scenario is that there is no impact of HSDPA on the layer that is

    already deployed. A Node B can handle HSDPA and non-HSDPA cells at the same

    time so there is no need for dedicated NodeB to HSDPA. However if no PA are added

    then HSDPA layer will use part of the current power capacity and this will reduce the

    coverage of the existing cells. Mobiles are spread over the two layers thanks to the

    Traffic Segmentation feature at RRC connection establishment, based on the release

    of the mobile and potentially on the traffic class indicated in the establishment cause.

    An HSDPA cell is not restricted to HSDPA services: it offers all UA4.2 services (on

    Cell_DCH and Cell_FACH) so there is no need to handover to the R99 layer to

    establish these services. However mobility between the two layers is managed

    through a cell reselection when the mobile is in an HSDPA call. This cell reslection is

    not based on quality criteria like in R99. Indeed, mobiles leaving the coverage of the

    HSDPA layer will lost his session from a UTRAN point of view but keep his PDP

    context from a core network point of view, so that it will perform a cell reselection on

    R99 layer.. It is also possible to configure a handover towards 2G (GPRS/EDGE)

    when alarm conditions are triggered but it will be a blind handover if the mobile is not

    able to perform measurements without compressed mode. For mobiles in Cell_DCH

    or Cell_FACH state, mobility is managed as usual even if they are on the HSDPA

    layer.

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    27/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 27/141

    Another possible scenario is to deploy HSDPA on a separate cluster (managed or not

    by a dedicated RNC) but in this case Traffic Segmentation cannot be used as it

    assumes a twin-cell topology (see 3.5.1 for more details on Traffic Segmentation).

    3.2.2 SINGLE CARRIER

    Even though deploying HSDPA on a separate layer is the preferred option, HSDPA

    can be configured on any cell and shared his resources with R99. The drawback of

    this scenario is that HSDPA traffic may impact R99 traffic by generating high

    interferences and may need to re-engineer the layer. If HSDPA is not deployed

    everywhere in the layer then an automatic channel type switching between DCH and

    HSDPA is performed when the UE enters in or leaves an HSDPA cell.

    3.3 HSDPA RESOURCES

    3.3.1 OVSF CODES

    3.3.1.1 CELL CONFIGURATION

    The following figure presents the cell configuration. This configuration provides up to

    15 SF16 codes for HS-PDSCH and up to 4 SF128 for HS-SCCH. The common

    channels (CPICH, P-CCPCH, S-CCPCH, ) take the equivalent of a SF32. All

    remaining OVSF codes can be used for non-HSDPA services (speech, multi-RABs..):

    Figure 12: Example of OVSF tree usage with HSDPA

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    28/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 28/141

    3.3.1.2 HSDPA CODE ALLOCATION

    In the OVSF code tree, all the common channels are allocated in the top of the tree,

    as illustrated by the figure below. In Nortel implementation, the HS-PDSCH SF16

    codes are allocated and reserved by the RNC at the bottom of the tree. Immediatelyabove, the HS-SCCH SF128 codes are allocated. These codes are allocated at cell

    setup and cannot be used or preempted for other services.

    SF = nSF = 4SF = 2SF = 1

    Cn,n-1

    Cn,0

    C4,3 = (1,-1,-1,1)

    C2,1 = (1,-1)

    C4,2 = (1,-1,1,-1)

    C1,0 = (1)

    C4,1 = (1,1,-1,-1)

    C2,0 = (1,1)

    C4,0 = (1,1,1,1)

    SF = nSF = 4SF = 2SF = 1

    Cn,n-1

    Cn,0

    C4,3 = (1,-1,-1,1)

    C2,1 = (1,-1)

    C4,2 = (1,-1,1,-1)

    C1,0 = (1)

    C4,1 = (1,1,-1,-1)

    C2,0 = (1,1)

    C4,0 = (1,1,1,1) Common channels

    n SF128 HS-SCCH

    m SF16 HS-PDSCH

    Figure 13: OVSF allocation strategy

    All the remaining codes are therefore contiguous and left for further DCH allocations.

    This includes associated DCH as well as any other calls mapped on DCH (e.g. speech

    calls, streaming, etc).

    Note that the maximum configuration (15 HS-PDSCH codes and 4 HS-SCCH codes)

    is not a valid one as it will leave no room in the OVSF tree for DCH (due to CCH

    occupancy) so it would not even be possible to allocate associated SRB for HSDPA

    calls.

    Note that OCNS code allocation is done before HSDPA configuration and it may lead

    to a conflict as HSDPA codes are always allocated from the bottom and need to be

    contiguous. In this case the HSDPA configuration will fail. The operator has to modify

    OCNS configuration to make it use non-conflicting codes.

    3.3.2 POWER

    In HSDPA, PA power has to be shared between common channels, DCH channels

    and HSDPA channels. In order to manage correctly this new power partition, some

    new margins or thresholds are added at RNC and NodeB level:

    At RNC level, a minimum power can be reserved for HSDPA.

    At NodeB level, a DCH margin allows to take into account the R99 power

    fluctuation due to power control.

    For more details, see section 287H6.4 288HPower Management.

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    29/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 29/141

    3.3.3 HSDPA CHANNELS & CQI

    3.3.3.1 PHYSICAL CHANNELS

    The following flowchart describes the timing relations between the different physical

    channels:

    HS-SCCH#2

    ACK ACK ACK

    7,5 slots

    HS-SCCH#1

    HS-PDSCH

    N_acknack_transmit = 2

    2 ms

    HS-DPCCH

    2 slots

    Figure 14: Timing relationship at NodeB between physical channels

    The mobile receives a HS-SCCH subframe (see the following figure) containing

    control information among which:

    Channelization-code-set information (7 bits slot #0 of subframe)

    Modulation scheme information (1 bit slot #0 of subframe), i.e.

    QPSK/16QAM

    Transport-block size information (6 bits slots #1 & #2 of subframe)

    Hybrid-ARQ process information (3 bits slots #1 & #2 of subframe)

    Redundancy and constellation version (3 bit slots #1 & #2 of subframe)

    New data indicator (1 bit slots #1 & #2 of subframe)

    UE identity (16 bits used as a mask in slots #0, #1 & #2 of subframe), i.e.

    subset of the HRNTI

    The SF is fixed to 128. It indicates to which UE data is intended to, on which codes

    and with which parameters. There are as many HS-SCCH transmitted during a TTI as

    scheduled user number.

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    30/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 30/141

    Data

    Slot #0 Slot #1 Slot #2

    1 HS-SCCH subframe = 2ms

    Tslot = 2560 chips = 40 bits

    Figure 15: HS-SCCH structure

    A mobile decoding its identity in the slot #0 of an HS-SCCH knows that it has been

    assigned resources on the HS-PDSCH channels (as indicated, with modulation, in this

    slot #0, other information are given in slots #1 and 2): the mobile receives a transportblock on one or several HS-PDSCH (see the following figure).

    M= 2 for QPSK

    (960 coded bits per TTI)

    M = 4 for 16QAM

    (1920 coded bits per TTI)

    Data

    Slot #0 Slot #1 Slot #2

    1 HS-PDSCH subframe = 2ms

    Tslot = 2560 chips = M*10*2k bits (k = 4, SF16)

    Figure 16: HS-PDSCH structure

    The HS-PDSCH on which is mapped the HS-DSCH carries only the data payload. The

    SF is equal to 16 and up to 15 codes can be reserved to HS-PDSCH per cell. One

    HS-DSCH can be mapped onto one or several HS-PDSCH (the maximum number of

    codes is given by UE capabilities).

    When addressed on HS-SCCH, the UE will then send feedback frame(s) on HS-

    DPCCH (SF = 256), roughly 7.5slots after HS-PDSCH frame, containing (see the

    following figure):

    The HARQ Ack/Nack;

    The CQI (Channel Quality Indication).

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    31/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 31/141

    CQI

    Subframe #0 Subframe #i Subframe #4

    1 radio frame = 10ms

    Tslot = 2560 chips

    = 10 bits

    ACK/NACK

    2.Tslot = 5120 chips

    = 20 bits

    Figure 17: HS-DPCCH structure

    The HARQ Ack is possibly repeated in consecutive HS-DPCCH subframes using the

    N_acknack_transmit parameter, as specified in [ 289HR5] 6A.1.1. The CQI is only sent in

    some specific subframes, as specified in [ 290HR5] 6A.1.1, depending on the following

    parameters:

    The CQI feedback cycle: k,

    the repetition factor of CQI: N_cqi_transmit.

    For more details on physical channel management, see [291HR6].

    3.3.3.2 CQI

    The HS-DPCCH channel contains the CQI computed by the mobile from P-CPICH

    power measurements. This CQI after demodulation and decoding by the NodeB

    (noted CQIreported) is not directly used by the scheduler. As shown by the following

    figure, the CQIreported undergoes two processing: a CQI averaging and a CQI

    adjustment based on BLER and DTX.

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    32/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 32/141

    HS-DPCCH demodulation

    and CQI decoding

    CQI averaging over several TTI

    in order to improve the reliability

    of this measurement.

    CQI provides to scheduler every TTI

    CQI adjustment based on BLER

    (to reach a BLER target)

    and DTX (in order to deactivate deficient ueby artificially setting its CQI to 0)

    CQIreported

    CQIaveraged

    CQIprocessed

    Figure 18: CQI Processing

    3.3.3.2.1 CQI AVERAGING

    An averaging over several TTI must be done on the CQI values (CQIreported) in order to

    improve the reliability of this measurement. The resulting value (CQIaveraged) should

    nevertheless be provided to entities needing this information (scheduler, etc) every

    TTI (the default CQI feedback cycle is equal to 2 ms). The chosen algorithm consists

    in a flexible averaging window which length will be based on CQI correlation

    estimation without noise level estimation (see [292HR8] for details).

    3.3.3.2.2 CQI ADJUSTMENT

    Two algorithms have been introduced to handle bad UE behaviors that would

    dramatically disrupt the system. Note that in the nominal case, these algorithmsshould not have any impact.

    The purpose of these algorithms is respectively to:

    Adjust the received CQI (CQIaveraged) in order to maintain an acceptable BLER

    on first transmission.

    Isolate a deficient UE which never responds (constant DTX detection).

    Both algorithms are processed just after the CQI averaging and can be processed inany order. The resulting CQI from both steps (referred as CQIprocessed in the document)

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    33/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 33/141

    constitutes the input of both flow control and scheduler algorithms (except HS-SCCH

    power control).

    3.3.3.2.3 CQI ADJUSTMENT ACCORDING TO BLER

    The first algorithm works on ACK/NACK statistics. The purpose is to correct some bias

    on the reported CQI that would lead to excessive BLER. Note that according to the

    specifications, the target on the first transmission when applying the reported CQI

    would be a 10% BLER. The idea is then to continuously compute the BLER and

    modify the reported CQI accordingly in order to reach such target.

    It then just consists in computing an offset to apply to the reported CQI (after

    averaging), referred in the following as CQI (CQIoutput = CQIaveraged + CQI). In case

    the input CQI equals 0, the offset doesnt apply even if positive and the CQI remains

    equal to 0. In case CQIoutput becomes inferior or equal to 0, the UE is not scheduled. In

    case CQIoutput becomes higher than 30, it is processed as a CQI 30.

    It is continuously updated with the following rules:

    A buffer of fixed size (= BufferSize) is created for each UE to compute the

    BLER.

    Anytime an ACK/NACK is received related to the 1st transmission of a

    transport block, the buffer is updated to store this information.

    o DTX is not taken into account in the buffer.

    o The feedback for retransmissions is not either taken into account.

    The buffer is filled in a circular manner (i.e. the new value replaces the oldest

    one when the buffer is full).

    When at least BufferSize stats have been received (the buffer is full), the

    number of NACK (NackNb) indication within the last BufferSize ones is

    computed. The offset is then updated according to the following rules:

    o If NackNb NackNbMin, the system is too good and bandwidth

    efficiency could be improved (throughput increase and/or powerreduction). CQI is increased by 1 and the buffer is reinitialized.

    o If NackNb > NackNbmax, the BLER is too high. Performances are

    then degraded. CQI is decreased by 1 and the buffer is reinitialized.

    o In all other cases, the system is considered in its stationary state and

    then behaves satisfactorily. CQI is not updated and the buffer is not

    reinitialised.

    Note that the buffer is only reinitialized when CQI is updated. This allows waiting for

    a certain time (BufferSize) before taking a new decision, and thus really evaluating theeffect of the new offset value. If the offset has not been updated, the buffer remains

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    34/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 34/141

    filled in a circular manner in order to react as soon as the situation changes (and not

    wait for a new period before identifying the problem). The offset is bounded and fits in

    the range [-30.. +30].

    This is illustrated by the following flowchart.

    ACK or NACK received (DTX are

    not taken into account

    End

    Y

    Isit the

    acknowledgement

    of a 1st

    transmission ?

    Update of statistics buffer

    (sliding window like)

    StatNb = min(StatNb +1, BufferSize)

    N

    CQI : offset to apply to averaged CQINackNb : number of Nack in the Stats buffer

    StatNb =

    BufferSize ?

    NackNb

    NackNbmin ?

    NackNb >

    NackNbmax ?CQI = CQI + 1

    CQI = CQI - 1No CQI update

    Reinit buffer

    StatNb = 0

    N

    N

    N

    Y

    Y

    Y

    Y

    Figure 19: CQI offset computation based on BLER

    The default values assumed in a first approach are:

    BufferSize = 100

    NackNbmin = 0

    NackNbmax = 30

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    35/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 35/141

    Tests must be done to tune these values. The higher the BufferSize the better the

    statistic but the longer the decision. A compromise must then be done between good

    precision (BufferSize high and NackNbmax/min close to 10%) and reactivity.

    3.3.3.2.4 DEFENSE BASED ON DTX DETECTION

    The second algorithm is more a defense mechanism against a deficient UE which

    would systematically not detect the HS-SCCH. The purpose is that upon detection of

    such UE (based on DTX statistic), it is deactivated by artificially setting its post-

    processed CQI to 0. That disables both the scheduler and the flow control algorithm.

    The UE would then be released by higher layers as its throughput equals 0.

    More in detail, the following steps are processed (see the following figure):

    As before, a buffer of fixed size (BufferSize) is created for each UE.

    Anytime a feedback information is received, the buffer is updated:

    o ACK, NACK and DTX are taken into account.

    o Feedback of all transmissions is considered (1st and retransmissions)

    as based on the HS-SCCH only.

    The buffer is also filled in a circular manner.

    As soon as the buffer is full, the evaluation may begin.

    If the buffer only contains DTX indications, the UE is considered as deficientand its CQI is set to 0. No recovery is then possible.

    If at least one ACK/NACK has been received in the last BufferSize

    transmissions, the UE is considered as valid and nothing is done.

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    36/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 36/141

    ACK/NACK/DTX received

    Permanently set the CQI ofcorresponding UE to 0

    (scheduler and flow control disable

    N

    Only DTX in

    the buffer ?

    Y

    Update of statistics buffer

    (sliding window like)StatNb = min(StatNb +1, BufferSize)

    StatNb =

    BufferSize ?

    End

    N

    Y

    Figure 20: Scheduler and Flow Control disabled

    Default value:

    BufferSize = 100

    Note: buffers of both algorithms are independent!

    3.4 UE CATEGORIES

    3GPP has standardized several UE categories to accommodate a large spectrum of

    HSDPA mobile implementations (See [293HR5]). The UE category provides the mobile

    capabilities like max number of HS-PDSCH codes supported, modulation schemes

    (16QAM, QPSK), MAC-HS transport block size, etc. Each UE category has a tablewith 30 CQI (channel quality indicator) values. Each CQI value provides complete

    information regarding the HS-DSCH to be received by UE in DL in the next TTI as

    shown below:

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    37/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 37/141

    Category 9

    Category 10

    Category 1-6

    Category 11-12

    15

    ...

    12

    10

    ...

    5

    ...

    5

    5

    4

    ...

    2

    2

    2

    1

    1

    1

    1

    1

    1

    Number of

    HS-PDSCH

    16-QAM0.89121602555830

    ...

    16-QAM0.7581601723726

    16-QAM0.7567201441125

    ...

    16-QAM0.753360716822

    ...

    16-QAM0.371660356516

    QPSK0.691440331915

    QPSK0.671120258314

    ...

    QPSK0.483209319*

    QPSK0.413207928

    QPSK0.341606507*

    QPSK0.481604616*

    QPSK0.391603775

    QPSK0.3303174*

    QPSK0.2402333*

    QPSK0.1801732*

    QPSK0.1401371*

    ModulationCode rateThroughput

    RLC level

    kb/s

    Transport Block

    SizeCQI value

    15

    ...

    12

    10

    ...

    5

    ...

    5

    5

    4

    ...

    2

    2

    2

    1

    1

    1

    1

    1

    1

    Number of

    HS-PDSCH

    16-QAM0.89121602555830

    ...

    16-QAM0.7581601723726

    16-QAM0.7567201441125

    ...

    16-QAM0.753360716822

    ...

    16-QAM0.371660356516

    QPSK0.691440331915

    QPSK0.671120258314

    ...

    QPSK0.483209319*

    QPSK0.413207928

    QPSK0.341606507*

    QPSK0.481604616*

    QPSK0.391603775

    QPSK0.3303174*

    QPSK0.2402333*

    QPSK0.1801732*

    QPSK0.1401371*

    ModulationCode rateThroughput

    RLC level

    kb/s

    Transport Block

    SizeCQI value

    Category 7-8

    Table 6: UE capabilities

    At the high end, UE category 10 can achieve a max RLC throughput = 12 Mbps using

    16QAM modulation and 15 OVSF codes SF16 (i.e. entire code tree). At the low end,

    UE category 12 achieves a max RLC throughput = 1.4 Mbps with QPSK modulation

    and using 5 OVSF codes SF16.

    3.5 CALL MANAGEMENT

    HSDPA only applies to the PS domain meaning that all the CS domain RABs are

    supported on dedicated channels.

    The Utran supports only the Traffic Classes Interactive & Background on HSDPA. So,

    the TC Streaming is served on DCH.

    Moreover, only the mono RB PS I/B are mapped on HSDPA implying that anycombination of RB PS I/B with RB CS call or with any other RB PS I/B (Multi PS) is

    served on DCH.

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    38/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 38/141

    RB configuration System behavior

    All RABs DCH available in HSDPA Cell

    Supported: no impact coming from HSDPA. All RAB

    combinations existing in this release are available on

    DCH in an HSDPA cell (either for a mobile that doesnot support HSDPA or for a RAB combination that is

    not supported on HSDPA)

    PS I/B on HSDPA Supported on HSDPA

    PS Streaming on HSDPA

    Speech on DCH + PS I/B on HSDPA

    CSD/DCH + PS I/B on HSDPA

    (PS I/B+PS I/B) on HSDPA

    (PS I/B+PS Streaming) on HSDPA

    Speech on DCH + (PS+PS) on HSDPA

    Not supported on HSDPA all channels are mapped

    on DCH

    Table 7: RB Configuration and system behaviour

    3.5.1 TRAFFIC SEGMENTATION

    In case HSDPA is deployed on one frequency layer on top of the R99 layer, it shall be

    possible to re-direct UEs on the proper frequency layer at call setup depending on its

    capabilities and requested establishment cause:

    an HSDPA capable UE camping on a R99 cell is re-directed to the HSDPA

    layer

    similarly, a R99 UE camping on a HSDPA cell can be re-directed to the R99

    layer.

    This feature impacts the choice of the target cell and the frequency layer at the call

    establishment.

    The main benefits are to allow HSDPA capable mobiles to benefit from HSDPA

    service and to avoid loading the HSDPA layer with R99 mobiles.

    3.5.1.1 TRAFFIC SEGMENTATION MECHANISM

    The redirection is performed by the RNC at the call setup phase based on the twin-

    cells configuration which is not compatible in this case with the f1/f2 mobility-capacity

    inter-frequency hard hand over.

    No redirection is performed by the RNC during the call meaning for example that

    even if a mobile, initially on HSDPA layer, reselects the R99 layer in AO cell_fach

    state for traffic resuming, it will be served using the DCH layer.

    Emergency calls are never redirected and are served on the layer selected by themobile to limit the probability of call setup failure.

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    39/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 39/141

    3.5.1.1.1 TRAFFIC SEGMENTATION CRITERIA:

    Two filtering can be operated in order to execute the redirection on the suitable layer:

    First criterion is based on the Access Stratum Release Indication IE in the

    RRC Connection Request message for the identification of the R99/R4

    mobiles versus the R5 mobiles

    Second optional criterion is based on the Establishment Cause IE in the RRC

    Connection Request message to distinguish the calls I/B potentially served on

    HSDPA from the others.

    This filtering is not necessarily representative of the services setup

    during the life of the RRC connection (for example the addition of a

    CS call on top of a hsdpa connection implies the switching of the PS

    connection on dch channel in the hsdpa layer).

    3.5.1.2 TRAFFIC SEGMENTATION PROCEDURE

    At the reception of the RRC Connection Request, the RNC identifies:

    the UE Release via the Access Stratum Release indicator IE knowing that

    R99/R4 mobiles dont support HSDPA configuration

    the requested service via the Establishment Cause IE knowing that traffic

    class Conversational and Streaming cant be served on HSDPA. This filteringis optional depending on the setting of the parameter

    isRedirectionBasedOnEstablishmentCause

    So, based on the information of Access Stratum Release indicator and Establishment

    Cause IE, the RNC can start the redirection procedure for the R5 mobiles requesting a

    PS I/B session.

    The redirection consists in indicating the target frequency in the RRC Connection

    Setup message via the Frequency Info IE, frequency corresponding to the one of the

    twin cell.

    The UE will send the RRC Connection Setup Complete towards the twin cell on the

    right layer.

    Hereafter the static mapping between the Establishment cause sent by the mobile in

    the RRC Connection Request and the suitable layer pointed by the RNC :

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    40/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 40/141

    Cause Suitable layer Meaning

    Originating Conversational Call DCH

    Originating Streaming Call DCH

    Originating Interactive Call HSDPA

    Originating Background Call HSDPA

    Originating Subscribed traffic Call HSDPA

    Request for access following a request forservice expressed by the mobile user

    Terminating Conversational Call DCH

    Terminating Streaming Call DCH

    Terminating Interactive Call HSDPA

    Terminating Background Call HSDPA

    Request for access, following a pagingindication received by the mobile.

    Emergency CallDCH or HSDPA

    (i.e. no re-direction)Speech emergency mobile originated call

    Inter-RAT cell re-selectionDCH or HSDPA

    (i.e. no re-direction)User location registration following an inter-

    system cell reselection

    Inter-RAT cell change orderDCH or HSDPA

    (i.e. no re-direction)

    User location registration following an inter-system cell change commanded by the

    source system

    Registration HSDPARequest for user registration, following a

    mobile power-on

    DetachDCH or HSDPA

    (i.e. no re-direction)Request for user de-registration, following a

    mobile power-off

    Originating High PrioritySignalling

    HSDPA

    Originating Low Priority

    Signalling

    DCH or HSDPA

    (i.e. no re-direction)

    Access request for signalling exchange, e.g.SMS,...

    Call re-establishment CS caseDCH or HSDPA

    (i.e. no re-direction)

    Access requested for service re-establishment (due to loss of radio

    connection)

    Call re-establishment PS caseDCH or HSDPA

    (i.e. no re-direction)Routing Area update due to "directed

    signalling call re-establishment" RRC release

    Terminating High PrioritySignalling

    DCH or HSDPA(i.e. no re-direction)

    Terminating Low PrioritySignalling

    DCH or HSDPA(i.e. no re-direction)

    Access request for network initiated signallingexchange, e.g. SMS, ...

    Terminating cause unknownDCH or HSDPA

    (i.e. no re-direction)This cause is received when no paging cause

    is provided from the Core Network.

    Table 8: RRC Connection Request and suitable layer

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    41/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 41/141

    3.5.1.2.1 PARAMETERS:

    isRedirectionForTrafficSegmentation : This parameter is used by the traffic

    segmentation feature, in order to redirect mobiles to the adequate layer according to

    the UE release indicator.

    Parameter isRedirectionForTrafficSegmentation Object InterFreqHhoFddCell

    Range & Unit Boolean{True, False}

    User Customer

    Class 3

    Granularity FDDCell

    Value

    isRedirectionBasedOnEstablishmentCause : This parameter is used by the traffic

    segmentation feature in order to take into the establishment cause if the redirection

    algorithm.

    Parameter isRedirectionBasedOnEstablishmentCause Object InterFreqHhoFddCell

    Range & Unit Boolean{True, False}

    User Customer

    Class 3

    Granularity FDDCell

    Value

    3.5.2 HSDPA CAC

    3.5.2.1 RAB MATCHING

    Any PS RAB request with Interactive or Background traffic class is matched to the

    HSDPA Radio Bearer configuration if the mobile is HSDPA capable and the primary

    cell of the active set supports HSDPA. If it is not the case, the request is mapped on

    DCH as usual (iRM CAC is performed).

    3.5.2.2 ADMISSION PHASE

    As today, this mechanism is triggered by the reception of RAB Assignment Request

    and follows the RAB matching process.

    In this implementation, the specific CAC admission process in the RNC for HSDPA is

    based on the number of simultaneous authorized users per cell to limit the

    degradation of the quality of service. So, iRM CAC is not played for HSDPA RAB.

    Any Interactive/background RAB request is admitted on HSDPA until the maximum

    number of simultaneous users allowed on HSDPA is reached for the cell. Unlike the

    iRM CAC performed for the RB mapped on DCH channels, the admission on hsdpa

    doesnt take into account any other criterion like RF power,

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    42/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 42/141

    In case the admission process fails, there is no fallback on DCH. The RAB request will

    be rejected.

    Once HSDPA CAC has been done, the RNC performs the admission on theassociated DCHs :

    Regarding DL SRB admission, CAC is performed as usual

    Regarding UL admission, CAC is performed as usual.

    3.5.2.2.1 PARAMETER

    maximumNumberOfUsers : Used for the HSDPA CAC done by the RNC. The

    number of users is per cell

    Parameter maximumNumberOfUsers Object HsdpaCellClass

    Range & Unit Integer[0..50]

    User customer

    Class 0

    Granularity RNC

    Value 20

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    43/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 43/141

    3.5.3 CALL SETUP (DATAFLOW)

    3.5.3.1 INITIAL CONNECTION PHASE:

    There is nothing specific to HSDPA in this phase.

    In the scenario of a mobile being redirected due to the traffic segmentation feature, the

    target frequency is indicated in the RRC Connection Setup (frequency info IE) but the

    call flow is the same.

    SGSNNode B RNCUE

    RRC/ RACH / RRC connection Request

    SCCP/ Connection Request (Initial UE msg (Activate PDP Context Request))

    SCCP/ Connection Confirm

    RRC/ FACH / RRC Connection Setup (DCCH, U-RNTI,

    RRC state = CELL_DCH, [frequency info])

    RRC/ RACH / RRC Connection Setup Complete

    RRC/ RACH / Initial Direct Transfer (Activate PDP Context Request, PS domain)

    The RRC Connection Setup message contains the

    signalling bearers (DCCH) definition. A UTRAN

    Radio Network Temporary Identity is also

    allocated to the UE.

    The target RRC state is set to CELL_DCH by the

    RNC

    If the mobile is redirected for traffic segmentation

    reason then frequency info is included

    In addition to the initial NAS message, the Initial Direct Transfer message

    also contains the CN domain identity, used by the RNC to route the NAS

    message to the relevant CN domain (i.e. CS or PS)

    The RRC Connection Request message initiated

    by the UE contains the establishment cause.

    The initial UE message is piggybacked in the

    SCCP connection resquest

    NBAP/ RL Setup Request)

    NBAP / RL Setup

    Figure 21: HSDPA Call setup / initial connection (Cell_DCH)

  • 7/27/2019 UMT IRC APP 016664 - HSDPA Engineering Guide v01.06 EXT Nortel

    44/141

    HSDPA Engineering Guide

    Nortel confidential

    UMT/IRC/APP/016664 01.06 / EN Preliminary 20/Jan/2006 Page 44/141

    3.5.3.2 RAB ALLOCATION PHASE:

    In this phase, only the NBAP Radio Link Reconfiguration procedure and RRC Radio

    Bearer Reconfiguration are modified because of HSDPA.

    SGSNNode B RNCUE

    SM/Activate PDP Context

    GMM/ Authentication and Ciphering

    GMM/ Authentication and Ciphering

    The UE is authenticated by the SGSN

    RANAP/ Security Mode Command

    RANAP/ Security Mode

    RRC/ Security Mode Command

    RRC/ Security Mode Complete

    The ciphering and integrity procedures are activa