High-Speed Downlink Packet Access (HSDPA) · Cellular Communication Systems Andreas Mitschele-Thiel, Jens Mueckenheim Nov. 2018 2 HSDPA Background Initial goals Establish a more spectral

High-Speed Downlink Packet Access (HSDPA)

HSDPA Background & Basics Principles: Adaptive Modulation & Coding, HARQ Channels/ UTRAN Architecture Resource Management: Fast Scheduling, Mobility Performance Results

Cellular Communication Systems 2 Andreas Mitschele-Thiel, Jens Mueckenheim Nov. 2018

HSDPA Background

Initial goals Establish a more spectral efficient way of using DL resources providing

data rates beyond 2 Mbps, (up to a maximum theoretical limit of 14.4 Mbps)

Optimize interactive & background packet data traffic, support streaming service

Design for low mobility environment, but not restricted Techniques compatible with advanced multi-antenna and receivers

Standardization started in June 2000

Broad forum of companies Major feature of Release 5

Enhancements in Rel.7 HSPA+

Advanced transmission to increase data throughput Signaling enhancements to save resources


HSDPA Basics

Evolution from R.99/ Rel.4 5 MHz BW Same spreading by OVSF and scrambling codes Turbo coding

New concepts in Rel.5 Adaptive modulation (QPSK vs. 16QAM), coding and multicodes

(fixed SF = 16) Fast scheduling in NodeB (TTI = 2 msec) Hybrid ARQ

Enhancements in Rel.7 HSPA+

Signaling enhancements 64QAM MIMO techniques, increase of the bandwidth


Higher Order Modulation

Standard modulation scheme in UMTS R.99 QPSK: 2 bit per symbol

With HSDPA the modulation can be switched between QPSK: 2 bit per Symbol 16QAM: 4 bit per Symbol

Low Bitrate → Robust High Bitrate → Sensitive to disturbances


Adaptive Modulation and Coding

For wireless data, link adaptation through Rate Control is more effective than Power Control.

Users in favorable channel conditions are assigned higher code rates and higher order modulation (16QAM). This means higher data rates = Reduced latency

Users in worse condition will get lower code rate and lower modulation

(QPSK). More robust transmission with lower data rate

Decision on modulation and coding is based on fast feedback from the UE

about channel quality (CQI).


Hybrid ARQ

H-ARQ automatically adapts to instantaneous channel conditions by: Fast retransmissions at MAC-sublayer Adding redundancy only when needed

The retransmitted packets are combined with the original packet to improve decoding probability.

Simple form of Hybrid ARQ shows significant gains over link adaptation alone.

Different schemes can be used for retransmission of the original data packet: Chase-Combining Incremental Redundancy


Fast Scheduling

Principle of fast scheduling Assign the resources to the best user(s) in time to maximise throughput Channels are uncorrelated → Multi-User Diversity Gain

There exist various variants, which mainly differ in their consideration of

channel quality and user throughput, e.g. Round Robin Max C/I Proportional Fair

With HSDPA the resource

management is shifted to the NodeB No Soft Handover


HS-DSCH Principle I

Channelization codes at a fixed spreading factor of SF = 16 Up to 15 codes in parallel

OVSF channelization code tree allocated by CRNC HSDPA codes autonomously managed by NodeB MAC-hs scheduler

Example: 12 consecutive codes reserved for HS-DSCH, starting at C16,4

Additionally, HS-SCCH codes with SF = 128 (number equal to simult. UE)

SF=8

SF=16

SF=4

SF=2

Physical channels (codes) to which HS-DSCH is mapped CPICH, etc.

C16,0 C16,15


HS-DSCH Principle II

Resource sharing in code as well as time domain:

Multi-code transmission, UE is assigned to multiple codes in the same TTI Multiple UEs may be assigned channelization codes in the same TTI

Example: 5 codes are reserved for HSDPA, 1 or 2 UEs are active within one TTI

Data to UE #1 Data to UE #2 Data to UE #3

Time (per TTI)

Code

not used


UMTS Channels with HSDPA

Cell 1

UE

Cell 2

R.99 DCH (in SHO) UL/DL signalling (DCCH) UL PS service UL/DL CS voice/ data

Rel.5 HS-DSCH DL PS service (Rel.6: DL DCCH)

= Serving HS-DSCH cell


HSDPA Channels

HS-PDSCH Carries the data traffic Fixed SF = 16; up to 15 parallel channels QPSK: 480 kbps/code, 16QAM: 960 kbps/code

HS-SCCH

Signals the configuration to be used next: HS-PDSCH codes, modulation format, TB information

Fixed SF = 128 Sent two slots (~1.3 msec) in advance of HS-PDSCH

HS-DPCCH

Feedbacks ACK/NACK and channel quality indicator (CQI) Fixed SF = 256, code multiplexed to UL DPCCH Feedback sent ~5 msec after received data


Timing Relations (DL)

NodeB Tx view Fixed time offset between the HS-SCCH information and the start of the

corresponding HS-DSCH TTI: τHS-DSCH-control (2 × Tslot= 1.33 msec) HS-DSCH and associated DL DPCH not time-aligned

TB size & HARQ Info

Downlink DPCH

HS-SCCH

DATA HS-PDSCH

3 × Tslot (2 msec)

HS-DSCH TTI = 3 × Tslot (2 msec)

τHS-DSCH-control = 2 × Tslot

Tslot (2560 chips)

ch. code & mod


Timing Relations (UL)

UE Rx view Alignment to m × 256 to preserve orthogonality to UL DPCCH HS-PDSCH and associated UL DPCH not time-aligned

(but “quasi synch”)

DATA

Uplink DPCCH

HS-PDSCH

HS-DPCCH

3 × Tslot (2ms)

m × 256 chips

τUEP = 7.5 × Tslot (5ms) 0-255 chips

Tslot (0.67 ms)

CQI A/N

CQI A/N

CQI A/N

CQI A/N


HSDPA Architecture

MAC-c/sh

MAC-d

RLC

RRC PDCP

Logical Channels

Transport Channels

MAC-b

BCH

BCCH DCCH DTCH

SRNC

CRNC

NodeB

DCH

Upper phy

DSCH FACH

Evolution from R.99/Rel.4

• HSDPA functionality is intended for transport of dedicated logical channels

• Takes into account the impact on R.99 networks

MAC-hs

HS-DSCH

HSDPA in Rel.5

• Additions in RRC to handle HSDPA

• RLC nearly unchanged (UM & AM)

• Modified MAC-d with link to MAC-hs entity

• New MAC-hs entity located in the NodeB

w/o

MAC

-c/s

h


MAC-hs in NodeB

MAC-hs

MAC – Control

HS-DSCH

Priority Queue distribution

MAC-d flows

Scheduling

Priority Queue

Priority Queue

Priority Queue

UE #1 UE #2

UE #N MAC-hs Functions Priority handling Flow Control

To RNC To UE

Scheduling Select which user/queue

to transmit Assign TFRC & Tx

power HARQ handling

Service measurements e.g. HSDPA provided

bitrate

TFRC: Transport Format and Resource Combination Cf. 25.321


MAC-hs in UE

MAC-hs

MAC – Control

Associated Uplink Signalling HS-DPCCH

To MAC-d

Associated Downlink Signalling HS-SCCH

HS-DSCH

HARQ

Reordering Reordering

Re-ordering queue distribution

Disassembly Disassembly

MAC-hs Functions HARQ handling

ACK/ NACK generation Reordering buffer handling

Associated to priority queues

Flow control per reordering buffer

Memory can be shared with AM RLC

Disassembly unit

Cf. 25.321


Data Flow through Layer 2

Higher Layer

L1

Higher Layer PDU

RLC SDU

MAC-d SDU

MAC-d PDU

…RLCheader

RLCheader…

MAC-d SDU

MAC-d PDU

CRC

……

MAC-dheader

MAC-dheader L2 MAC-d

(non-transparent)

L2 RLC(non-transparent)Segmentation

&Concatenation

Reassembly

Higher Layer PDU

RLC SDU

MAC-hs SDU MAC-hs SDU…MAC-hsheader L2 MAC-hs

(non-transparent)Transport Block (MAC-hs PDU)

…


Hybrid Automatic Repeat Request

HARQ is a stop-and-wait ARQ Up to 8 HARQ processes per UE

Retransmissions are done at MAC-hs layer, i.e. in the Node B Triggered by NACKs sent on the HS-DPCCH

The mother code is a R = 1/3 Turbo code Code rate adaptation done via rate matching, i.e. by puncturing and

repeating bits of the encoded data Two types of retransmission

Incremental Redundancy Additional parity bits are sent when decoding errors occured Gain due to reducing the code rate

Chase Combining The same bits are retransmitted when decoding errors occured Gain due to maximum ratio combining

HSDPA uses a mixture of both types


HARQ Processes

HARQ is a simple stop-and-wait ARQ Example

RTTmin = 5 TTI Synchronous retransmissions (MAC-hs decides on transmission)

UE support up to 8 HARQ processes (configured by NodeB) Min. number: to support continuous reception Max. number: limit of HARQ soft buffer Number of HARQ processes configured specifically for each UE category

1

1

2 2

2

3 4 5 3

3 4 5

1

RTTHARQ

Data HS-PDSCH

ACK/NACK HS-DPCCH


HSDPA UE Categories

The specification allows some freedom to the UE vendors

12 different UE categories for HSDPA with different capabilities (Rel.5)

The UE capabilities differ in Max. transport block size (data rate) Max. number of codes per HS-DSCH Modulation alphabet (QPSK only) Inter TTI distance (no decoding of HS-DSCH in each TTI) Soft buffer size

The MAC-hs scheduler needs to take these restrictions into account


HSDPA – UE Physical Layer Capabilities

HS-DSCH Category

Maximum number of HS-DSCH

multi-codes

Minimum inter-TTI interval

Maximum MAC-hs TB size

Total number of soft channel

bits

Theoretical maximum data rate (Mbit/s)

Category 1 5 3 7298 19200 1.2

Category 2 5 3 7298 28800 1.2

Category 3 5 2 7298 28800 1.8

Category 4 5 2 7298 38400 1.8

Category 5 5 1 7298 57600 3.6

Category 6 5 1 7298 67200 3.6

Category 7 10 1 14411 115200 7.2

Category 8 10 1 14411 134400 7.2

Category 9 15 1 20251 172800 10.1

Category 10 15 1 27952 172800 14.0

Category 11* 5 2 3630 14400 0.9

Category 12* 5 1 3630 28800 1.8

cf. TS 25.306 Note: UEs of Categories 11 and 12 support QPSK only


Channel Quality Indicator (CQI)

Signaled to the Node B in UL each 2 msec on HS-DPCCH

Integer number from 0 to 30 corresponds to a Transport Format Resource Combination (TFRC) given by Modulation Number of channelization codes Transport block size

For the given conditions the BLER for this TFRC shall not exceed 10%

Mapping defined in TS 25.214 for each UE category


CQI – Mapping Table

CQI value Transport Block Size

Number of HS-PDSCH Modulation

Reference power adjustment ∆

NIR XRV

0 N/A Out of range

1 137 1 QPSK 0

…

6 461 1 QPSK 0

7 650 2 QPSK 0

…

15 3319 5 QPSK 0

16 3565 5 16-QAM 0

…

23 9719 7 16-QAM 0

24 11418 8 16-QAM 0

25 14411 10 16-QAM 0

26 17237 12 16-QAM 0

27 21754 15 16-QAM 0

28 23370 15 16-QAM 0

29 24222 15 16-QAM 0

30 25558 15 16-QAM 0

28800 0

Tables specified in TS 25.214 For each UE category Condition:

BLER ≤ 10% Example is for

UE category 10


3G (R.99) with dedicated channels

C/I C/I

C/I

CQI

CQI

CQI

2 TTI @1.2M

2 TTI @760k

7 TTI @614k

1 TTI @1.2M

64k 64k

64k

Note: No fast channel quality feedback

3G with high speed feedback/scheduling on shared channels

HSDPA Fast Scheduling


HSDPA Resource Allocation

Scheduler

QoS & Subscriber Profile QoS: guar. bitrate, max. delay

GoS: gold/ silver/ bronze

Feedback from UL CQI, ACK/NACK

UE capabilities max. TFRC

Radio resources Power, OVSF codes

UE service metrics Throughput, Buffer Status

Scheduler Output • Scheduled Users • TFRC: Mod., TB size, # codes, etc. • HS-PDSCH power

• Scheduling targets − Maximize network throughput − Satisfy QoS/ GoS constraints − Maintain fairness across UEs and traffic streams


Scheduling Disciplines

Task

Select UEs (and associated priority queues) to transmit within next TTI Usually this is done by means of ranking lists

Depending on the ranking criterion it can be distinguished between three major categories Round Robin: allocate each user equal amount of time Proportional Fair: equalise the actual channel rate / throughput ratio Max C/I: prefer the users with good channel conditions

To provide GoS/ QoS additional inputs are to be used Additional scheduling weights and rate constraints based on the

requested GoS/ QoS This can be traded-off with channel conditions Special scheduling schemes are needed for providing delay critical

services, e.g. VoIP


Comparison of Schedulers

Simple Round Robin doesn’t care about actual channel rate Proportional Fair offers higher cell throughput QoS aware algorithm offers significantly higher user perceived throughput than

PF with similar cell throughput

aggregated cell throughput

0

500

1000

1500

2000

2500

Round Robin Proportional Fair QoS aw are

aver

age

thro

ughp

ut [k

bps]

user perceived throughput

0%

20%

40%

60%

80%

100%

0 100 200 300 400 500 600average throughput [kbps]

Perc

enta

ge o

f use

rs

rece

ivin

g th

roug

hput

Round Robin

Proportional Fair

QoS aw are


Mobility Procedures I

HS-DSCH for a given UE belongs to only one of the radio links assigned to the UE (serving HS-DSCH cell)

The UE uses soft handover for the uplink, the downlink DCCH and any simultaneous CS voice or data

Using existing triggers and procedures for the active set update (events 1A, 1B, 1C)

Hard handover for the HS-DSCH, i.e. Change of Serving HS-DSCH Cell within active set

Using RRC procedures, which are triggered by event 1D


Mobility Procedures II

Inter-NodeB serving HS-DSCH cell change Note: MAC-hs needs to be transferred to new NodeB !

NodeB

NodeB

MAC-hs

NodeB

NodeB

MAC-hs

Source HS-DSCH Node B

Target HS-DSCH Node B

Serving HS-DSCH radio link

Serving HS-DSCH radio link

s t

SRNC SRNC


HS-DSCH Serving Cell Change

Event 1D: change of best cell within the active set Hysteresis and time to trigger to avoid ping-pong

(HS-DSCH: 1…2 dB, 0.5 sec)

Reporting event 1D

Measurement quantity

Time

CPICH 2

CPICH 1

CPICH3

Hysteresis

Time to trigger


Handover Procedure

Example: HS-DSCH hard handover (synchronized serving cell change)

SRNC =

DRNC Target

HS-DSCH cell UE

RL Reconfiguration Prepare RL Reconfiguration Ready

Transport Channel Reconfiguration

Transport Channel Reconfiguration Complete

Source HS-DSCH cell

ALCAP Iub HS-DSCH Data Transport Bearer Setup If new NodeB

Synchronous Reconfiguration with Tactivation RL Reconfiguration Commit

ALCAP Iub HS-DSCH Data Transport Bearer Release

DATA

Reset MAC-hs entity

Serving HS-DSCH cell change decision i.e. event 1D

RL Reconfiguration Prepare

RL Reconfiguration Ready

RL Reconfiguration Commit


HSDPA – Managed Resources

a) OVSF Code Tree

b) Transmit Power

SF=8

SF=16

SF=4

SF=2

Codes reserved for HS-PDSCH/ HS-SCCH

C16,0 C16,15

Codes available for DCH/ common channels

Border adjusted by CRNC

Tx power available for HS-PDSCH/ HS-SCCH Tx power available for DCH/ common channels

Border adjusted by CRNC

Note: CRNC assigns resources to Node B on a cell basis


Cell and User Throughput vs. Load

36 cells network UMTS composite channel model FTP traffic model (2 Mbyte

download, 30 sec thinking time)

The user throughput is decreased when increasing load due to the reduced service time

The cell throughput increases with the load because overall more bytes are transferred in the same time

0

500

1000

1500

2000

2500

4 6 8 10 12 14 16 18

Thro

ughp

ut [

kbit

/s]

Number of Users/ Cell

Load Impact

Mean User Throughput

Aggregated Cell Throughput


HSDPA Performance per Category

36 cells network UMTS composite channel model FTP traffic model (2 Mbyte

download, 30 sec thinking time)

Higher category offers higher max. throughput limit Cat.6: 3.6 MBit/sec Cat.8: 7.2 MBit/sec

Max. user perceived performance

increased at low loading Cell performance slightly better

Cat 6 - Cat 8 Comparison

0

500

1000

1500

2000

2500

Cat 6/ 10 users Cat 8/ 10 users Cat 6/ 20 users Cat 8/ 20 users

thro

ughp

ut (k

bps)

Mean User ThroughputPeak User ThroughputAggregated Cell Throughput


Impact from Higher Layers

Maximum MAC-hs throughput is determined by the MAC-d PDU size and the max. number of MAC-d PDUs, which fit into the max. MAC-hs PDU

Maximum RLC throughput is further limited by The RLC window size, which

is limited to 2047 PDUs Round-trip time RTT

0

2000

4000

6000

8000

10000

12000

14000

Cat.6 – 336bit Cat.8 – 336bit Cat.8 – 656bit Cat.10 – 336bit Cat.10 – 656bit

Thro

ughp

ut [

kbit

/s]

Higher Layer Impact

Max. RLC Throughput, RTT = 120msec

Max. RLC Throughput, RTT = 80msec

Max. MAC-hs Throughput


Coverage Prediction with HSDPA

Example Scenario 15 users/cell Pedestrian A channel

model Plot generated with field

prediction tool

HSDPA Throughput depends on location

High Speed Uplink Packet Access (HSUPA)

Enhanced Uplink Dedicated Channel (EDCH)

EDCH Background & Basics Channels/ UTRAN Architecture Resource Management: Scheduling, Handover Performance Results


Background

E-DCH is a Rel.6 feature with following targets Improve coverage and throughput, and reduce delay of the uplink

dedicated transport channels Priority given to services such as streaming, interactive and background

services, conversational (e.g. VoIP) also to be considered Full mobility support with optimizing for low/ medium speed Simple implementation Special focus on co-working with HSDPA

Standardization started in September 2002

Study item completed in February 2004 Stage II/ III started in September/ December 2004 Release 6 frozen in December 2005/ March 2006 Various improvements have been introduced in Rel.7 & Rel.8


E-DCH Basics

E-DCH is a modification of DCH – Not a shared channel, such as HSDPA in the downlink !!

PHY taken from R.99 Turbo coding and BPSK modulation Power Control 10 msec/ 2 msec TTI Spreading on separate OVSF code, i.e. code mux with existing PHY

channels

MAC similarities to HSDPA Fast scheduling Stop and Wait HARQ: but synchronous

New principles

Intra Node B “softer” and Inter Node B “soft” HO supported for the E-DCH with HARQ

Scheduling distributed between UE and NodeB


E-DCH Scheduling

UE sends scheduling information MAC-e signaling On E-DPCCH: “happy bit”

NodeB allocates the resources Absolute/ relative scheduling grants Algorithms left open from standards

Depending on the received grants, UE decides on transmission Maintains allocated resources by means of internal serving grants Selects at each TTI amount of E-DCH data to transmit Algorithms fully specified by UMTS standard

DATA

UE NodeB

UE detects data in buffer

Scheduling information Scheduler takes UE for scheduling

Scheduling grant

Scheduling information

Scheduling grant

Scheduling grant


UMTS Channels with E-DCH

Cell 1

UE

Cell 2

R.99 DCH (in SHO) UL/DL signalling (DCCH) UL/DL CS voice/ data

Rel.5 HS-DSCH (not shown) DL PS service (DTCH) DL signalling (Rel.6, DCCH)

Rel.6 E-DCH (in SHO) UL PS service (DTCH) UL Signalling (DCCH)

= Serving E-DCH cell


E-DCH Channels

E-DPDCH Carries the data traffic Variable SF = 256 … 2 UE supports up to 4 E-DPDCH

E-DPCCH Contains the configuration as used on E-DPDCH Fixed SF = 256

E-RGCH/ E-HICH

E-HICH carries the HARQ acknowledgements E-RGCH carries the relative scheduling grants Fixed SF = 128 Up to 40 users multiplexed onto the same channel by using specific

signatures E-AGCH

Carries the absolute scheduling grants Fixed SF = 256


Timing Relation (UL)

E-DPDCH/ E-DPCCH time-aligned to UL DPCCH

Uplink DPCCH

Subframe #0

E-DPDCH/ E-DPCCH

3 × Tslot (2 msec)

Subframe #1 Subframe #2 Subframe #3 Subframe #4

10 msec

CFN

15 × Tslot (10 msec)

CFN+1

0.4 × Tslot (1024 chips) ±148chips

Downlink DPCH

10 msec TTI 2msec TTI

CFN


HSUPA UE Categories

Note: When 4 codes are transmitted, 2 codes are transmitted with SF2 and 2 with SF4

E-DCH Category

Max. num. Codes

Min SF EDCH TTI Maximum MAC-e TB size

Theoretical maximum PHY data rate (Mbit/s)

Category 1 1 SF4 10 msec 7110 0.71

Category 2 2 SF4 10 msec/ 2 msec

14484/ 2798

1.45/ 1.4



20000/ 5772

2.0/ 2.89



20000/ 11484

2.0/ 5.74

cf. TS 25.306


E-DCH UTRAN Architecture

MAC-c/sh

MAC-d

RLC

RRC PDCP

Logical Channels

Transport Channels

MAC-b

BCH

BCCH DCCH DTCH

SRNC

CRNC

NodeB

DCH

Upper phy

DSCH FACH

Evolution from Rel.5

• E-DCH functionality is intended for transport of dedicated logical channels (DTCH/ DCCH)

MAC-hs

HS-DSCH

w/o

MAC

-c/s

h

MAC-d flows

MAC-e

MAC-es MAC-d flows

EDCH

E-DCH in Rel.6

• Additions in RRC to configure E-DCH

• RLC unchanged (UM & AM)

• New MAC-es entity with link to MAC-d

• New MAC-e entity located in the NodeB

• MAC-e entities from multiple NodeB may serve one UE (soft HO)


MAC-e/es in UE

MAC-e/es Functions Priority handling

Per logical channel

Multiplexing MAC-d flow concept Mux of data from multiple

MAC-d flows into single MAC-e PDU

Scheduling

Maintain scheduling grant E-TFC selection HARQ handling

Cf. 25.309

MAC-es/e

MAC – Control

Associated Uplink Signalling: E-TFCI, RSN, happy bit

(E-DPCCH)

To MAC-d

HARQ

Multiplexing E-TFC Selection

Associated Scheduling Downlink Signalling

(E-AGCH / E-RGCH(s))

Associated ACK/NACK signaling (E-HICH)

UL data (E-DPDCH)


MAC-e in NodeB

MAC-e Functions Per user

HARQ handling: ACK/ NACK generation

De-multiplexing E-DCH control:

Rx/ Tx control signals

E-DCH scheduling for

all users Assign resources

(scheduling grants)

Iub overload control Cf. 25.309

MAC-e

MAC – Control

E-DCH Associated Downlink Signalling

Associated Uplink

Signalling

MAC-d Flows

De-multiplexing

HARQ entity

E-DCH Control

E-DCH Scheduling

Common RG

UE #2

UE #N

UE #1


MAC-es in SRNC

MAC-es

MAC – Control

From MAC-e in NodeB #1

To MAC-d

Disassembly

Reordering Queue Distribution

Reordering Queue Distribution

Disassembly

Reordering/

Combining

Disassembly

Reordering/ Combining

Reordering/ Combining

From MAC-e in NodeB #k

MAC-d flow #1 MAC-d flow #n

MAC-es Functions

Queue distribution

Reordering

Per logical channel

In-sequence delivery

Macro-diversity combining: frame selection

Disassembly

Cf. 25.309


Data Flow through Layer 2

MAC-es PDU MAC-e header

DATA Header

DATA

DATA

DDI N Padding (Opt)

RLC PDU:

MAC-e PDU:

DDI N DATA

MAC-d PDU:

DDI

RLC

MAC-d

MAC-e/es

PHY

TSN DATA DATA MAC-es PDU:

DATA

DDI: Data Description Indicator (6 bit)

MAC-d PDU size

Log. Channel ID

Mac-d flow ID

N: Number of MAC-d PDUs (6 bit)

TSN: Transmission Sequence Number (6 bit)


Hybrid ARQ Operation

N-channel parallel HARQ with stop-and-wait protocol Number of HARQ processes N to allow uninterrupted E-DCH transmission

10 msec TTI: 4 2 msec TTI: 8

Synchronous retransmissions Retransmission of a MAC-e PDU follows its previous HARQ (re)transmission

after N TTI = 1 RTT Incremental Redundancy via rate matching

Max. # HARQ retransmissions specified in HARQ profile

New Tx 2 New Tx 3 New Tx 4 Re-Tx 1 New Tx 2 Re-Tx 3 New Tx 4 Re-Tx 1 Re-Tx 2 New Tx 1

ACK

ACK

NACK

NACK

NACK

NACK


E-DCH UE Scheduling

UE maintains internal serving grant SG SG are quantized Maximum E-DPDCH/ DPCCH power ratio (TPR),

which are defined by 3GPP Reception of absolute grant: SG = AG

No transmission: SG = “Zero_Grant” Reception of relative grants: increment/ decrement index of SG in the

SG table AG and RG from serving RLS can be activated for specific HARQ

processes for 2 msec TTI UE selects E-TFC at each TTI Allocates the E-TFC according to the given restrictions

Serving grant SG UE transmit power

Provides priority between the different logical channels


Scheduling Grant Table

Index Scheduled Grant

37 (168/15)2*6 36 (150/15)2*6 35 (168/15)2*4 34 (150/15)2*4 33 (134/15)2*4 32 (119/15)2*4 31 (150/15)2*2 30 (95/15)2*4 29 (168/15)2

• • •

14 (30/15)2 13 (27/15)2 12 (24/15)2 11 (21/15)2 10 (19/15)2 9 (17/15)2 8 (15/15)2 7 (13/15)2 6 (12/15)2 5 (11/15)2 4 (9/15)2 3 (8/15)2 2 (7/15)2 1 (6/15)2 0 (5/15)2

Scheduling grants are max. E-DPDCH/ DPCCH power ratio (TPR – traffic to pilot ratio) Power Ratio is related to UE data

rate

Relative Grants SG moves up/ down when RG = UP/

DOWN

Absolute Grants SG jumps to entry for AG 2 reserved values for ZERO_GRANT/

INACTIVE


Timing Relation for Scheduling Grants

AG and RG associated with specific uplink E-DCH TTI, i.e. specific HARQ process Association based on the timing of the E-AGCH and E-RGCH.

Timing is tight enough that this relationship is un-ambiguous. Example: 10 msec TTI

1 2 3 4 1 2 3

E-RGCH E-AGCH

E-DCH

HARQ process number

Scheduling decision

Load estimation, etc

• AG applied to this HARQ process

• RG interpreted relative to the previous TTI in this HARQ process.


Scheduling Information

Happy bit signaling One bit status flag send on E-DPCCH at each TTI Criterion for happy bit

Set to “unhappy” if UE is able to send more data than given with existing serving grant

Otherwise set to “happy” Scheduling Information Reporting Content of MAC-e report

Provides more detailed information (log. channel, buffer status, UE power headroom)

Will be sent less frequently (e.g. every 100 msec) Parameters adjusted by RRC (e.g. reporting intervals, channels to

report)


HSUPA Scheduling

EDCH NodeB Scheduler

QoS Parameters Throughput bounds

Feedback from UE Scheduling Information

Reports

Other constraints NodeB decoding capabilities

Iub bandwidth limit

UE capabilities

Radio resources UL Load (interference)

Allocate (absolute/ relative) Scheduling Grants (max. allowed power offsets)

UE allocates transport formats according to the allocated grants


NodeB Load Scheduling Principle

E-DCH scheduler constraint Keep UL load within the limit

Scheduler controls: E-DCH load portion of non-serving

users from other cells E-DCH resources of each serving user

of own cell Principles:

Rate vs. time scheduling Dedicated control for serving users Common control for non-serving

users

Note: Scheduler cannot exploit fast fading !

Non E-DCH

Non-serving E-DCH users

Serving E-DCH users

UL Load UL Load target

UE #1

UE #m


Rate Scheduling

UEs are continuously active Data rate is incrementally increased/

decreased by relative scheduling grants No synch between UEs required Load variations can be kept low For low to medium data rates

Time Scheduling

UEs are switched on/ off by absolute scheduling grants

UEs should be in synch Load variations might be large For (verry) high data rates

E-DCH Scheduling Options

time

rate

time

rate

UE1 UE2 UE3 UE1

UE1

UE2

UE3


Non-scheduled Mode

Configured by the SRNC

UE is allowed to send E-DCH data at any time Signaling overhead and scheduling delay are minimized

Support of QoS traffic on E-DCH, e.g. VoIP & SRB Characteristics

Resource given by SRNC: Non-scheduled Grant = max. # of bits that can be included in a MAC-e PDU UTRAN can reserve HARQ processes for non-scheduled transmission

Non-scheduled transmissions defined per MAC-d flow Multiple non-scheduled MAC-d flows may be configured in parallel One specific non-scheduled MAC-d flow can only transmit up to the non-

scheduled grant configured for that MAC-d flow

Scheduled grants will be considered on top of non-scheduled transmissions Scheduled logical channels cannot use non-scheduled grant Non-scheduled logical channels cannot transmit data using Scheduling Grant


E-DCH Power Control – Tx Power of E-DPCCH/ E-DPDCH

E-DCH is power-controlled the same way as R.99 DCH E-DPCCH/ E-DPDCH power controlled with offsets relative to DPCCH

DPCCH still under closed inner/ outer loop power control E-DPCCH/ DPCCH offset signaled by RRC

E-DPDCH/ DPCCH offset adjusted according to selected E-TFC Reference PO/ reference E-TFCI signaled by RRC Calculated for other E-TFCI from reference PO (specified in standard) Additional offset ∆HARQ in HARQ profile for each MAC-d flow to satisfy

different QoS requirements

E-DCH quality control loop

Each MAC-es PDU received by the SRNC contains indication of how many retransmissions were required to decode it Measure of the received E-DCH quality

SRNC may react as follows Update SIR target setting for DPCCH via DCH FP signaling Signal new power offset settings to UE/ NodeB via RRC signaling


E-DCH Operation in Soft Handover

Macro-diversity operation on multiple NodeBs Softer handover combining in the same NodeB Soft handover combining in RNC (part of MAC-es)

Independent MAC-e processing in both NodeBs HARQ handling rule: if at least one NodeB tells ACK, then ACK Scheduling rule: relative grants “DOWN” from any NodeB have

precedence

NodeB 1 NodeB 2

UE

scheduling grant HARQ ACK/ NACK

scheduling grant HARQ ACK/ NACK


Mobility Handling

The UE uses soft handover for associated DCH as well as for E-DCH

Using existing triggers and procedures for the active set update (events 1A, 1B, 1C)

E-DCH active set is equal or smaller than DCH active set New event 1J: non-active E-DCH link becomes better than active one

The UE receives AG on E-AGCH from only one cell out of the E-DCH

active set (serving E-DCH cell) E-DCH and HSDPA serving cell must be the same Hard Handover, i.e. change of serving E-DCH cell Using RRC procedures, which maybe triggered by event 1D

Could also be combined with Active Set Update


Mobility Procedures

Inter-NodeB serving E-DCH cell change within E-DCH active set Note: MAC-e still established in both NodeBs !

NodeB

NodeB

MAC-e

NodeB

NodeB

MAC-e

Serving E-DCH radio link

Serving E-DCH radio link

s t

SRNC SRNC

MAC-es MAC-es

MAC-e MAC-e


Serving E-DCH Cell Change

Handover of E-DCH scheduler control No changes in UL transport bearer No MAC-es RESET

Handover of HS-DSCH serving cell DL transport bearer setup MAC-hs RESET

SRNC =

DRNC Target serving

E-DCH cell UE

RL Reconfiguration Prepare RL Reconfiguration Ready

Transport Channel Reconfiguration

Transport Channel Reconfiguration Complete

Source serving E-DCH cell

If new NodeB

Synchronous Reconfiguration with Tactivation RL Reconfiguration Commit

Serving E-DCH cell change decision i.e. event 1D

RL Reconfiguration Prepare

RL Reconfiguration Ready

RL Reconfiguration Commit

UE receives now AG & dedicated RG from target cell


E-DCH RRM Principle

E-DCH resources controlled by UL load target E-DCH non-serving load portion

NodeB schedules E-DCH users

according to RNC settings Priority for non E-DCH traffic

RNC still controls non E-DCH load

portion By means of e.g. admission/

congestion control Based on an estimate of non-

EDCH load Non E-DCH

Non-serving E-DCH users

Serving E-DCH users

UL Load UL Load target

Non E-DCH load portion


User Throughput vs. Aggregate Cell Throughput

36 cells network UMTS composite channel

model FTP traffic model (2 Mbyte

upload, 30 seconds thinking time)

Maximum cell throughput reached for about 7…8 UEs per cell Cell throughput drops if #UEs

increases further since the associated signaling channel consume UL resources too

#UEs/cell 1

2

3

4

5 6 7 8

9 10

1 UE/ cell

2

3

4

5678

10121416182022

240

200

400

600

800

1000

1200

1400

1600

1800

2000

200 400 600 800 1000 1200 1400 1600 1800

Aver

age

Use

r Thr

ough

put [

kbps

]

Aggregate Cell Throughput [kbps]

10ms TTI 2ms TTI


Single User Performance

Average user throughput (RLC layer) for different channel profiles 1 UE in the network

1 target HARQ transmission

For AWGN channel conditions: 10 ms TTI: up to 1.7 Mbps

(near theoretical limit of 1.88 Mbps)

2 ms TTI: up to 3.6 Mbps (below theoretical limit of 5.44 Mbps)

E.g. due to restrictions from RLC layer (window size, PDU size)

0

500

1000

1500

2000

2500

3000

3500

4000

AWGN PedA3 PedA30 VehA30 VehA120

Scenario

Ave

rage

Use

r Thr

ough

put [

kbps

]

2ms TTI 10ms TTI

Enhanced High-Speed Packet Access HSPA+

Background: HSPA Evolution Higher Data Rates Signaling Improvements Architecture Evolution/ Home NodeB


HSPA+

The evolution of UMTS HSPA Corresponding to UMTS Release 7 and beyond

Motivation of HSPA+ 3GPP Long Term Evolution (LTE) being rolled-out, but not backwards

compatible with HSPA Investment protection needed for current HSPA deployments

Main goals Performance and flexibility comparable to LTE in 5 MHz Optimized packet-only mode for voice and data Backward compatible with Release 99 through Release 6 Smooth migration path to LTE through commonality and facilitate

joint technology operation Requiring simple infrastructure upgrade from HSPA to HSPA+

HSPA+ defines a broad framework and set of requirements Improvement of the radio interface Architecture evolution


Higher Order Modulations (HOMs)

BPSK 1 bit/symbol

16QAM 4 bits/symbol

64QAM 6 bits/symbol

Uplink Downlink

Increases the peak data rate in a high SNR environment Very effective for micro cell and indoor deployments


Peak Rate Performance Benefits of HSPA+

0

5

10

15

20

25

30

35

40

45

50

Pea

k Th

roug

hput

in M

bps

14

21

28

42

5.7

11.5

HSDPA

HSPA+ 64QAM

HSPA+ 16QAM

2x2 MIMO

HSPA+ 64QAM*

2x2 MIMO

HSUPA

HSPA+ 16QAM

Downlink

Uplink

*Release 8

Uplink and Downlink peak rates similar to LTE peak rates in 5 MHz

Major increase HSPA peak

rates by Higher Order Modulations

Data rate benefits for users in ideal channel conditions (e.g. static users, fixed users close to the cell center, lightly loaded conditions)


HSDPA Performance with 64QAM

Without 64QAM With 64QAM Gain

Cell Throughput 6.9 Mbit/s 7.65 Mbit/s 10.7%

95%-tile User Throughput 7.1 Mbit/s 8.7 Mbit/s 22.5%

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

Cat 10/ 15 users Cat 14/ 15 users

thro

ughp

ut/ k

bps

average user throughput 95% user throughput ave. cell throughput

Single micro-cell scenario, advanced receivers required


16QAM for E-DCH

16QAM specified in the uplink for HSPA Evolution, for use with the 2 msec

TTI and with 4 multicodes (2xSF2 + 2xSF4) Increases peak rate from 5.76 Mbps to 11.52 Mbps

Performance results showed: 16QAM requires very good radio conditions Enhancement of the radio architecture needed (transmitter, receiver)

BPSK

BPSK

BPSK

BPSK

SF2

SF2

SF4

SF4

I

Q

I

Q

SF2

SF4

16 QAM

I

Q

16 QAM

I

Q


Basic MIMO Channel

The HSDPA MIMO channel consists of 2 Tx and 2 Rx antennas Each Tx antenna transmits a different signal The signal from Tx antenna j is received at all Rx antennas i Channel capacity can be increased by up to a factor of two

Coding/Modulation/ Weighting/Mapping

Weighting/Demapping Demodulation/Decoding

Tx Rx

H

HVUH Λ=


MIMO in HSPA+

Release 7 MIMO for HSDPA (D-TxAA) 2 x 2 MIMO scheme The mobile reports the rank of the channel and the preferred precoding weights

periodically (PCI) Dynamic switching between single stream and dual stream transmission is

supported by the NodeB scheduler

Weight Generation

w 1 w 4

Determine weight info message from the uplink

w 2 w 3

TrCH processing

HS-DSCH TrCH processing

HS-DSCH

Spread/scramble

Primary transport block

Primary: Always present for scheduled UE

Secondary: Optionally present for scheduled UE

Secondary transport block

∑

Ant 1

Ant 2

∑

CPICH 1

CPICH 2

w 1

w 2

w 3

w 4

∑

∑

3 1 4 2 21 1 1 11/ 2, , , , ,

2 2 2 2j j j jw w w w w + − − + − − = = = − ∈


MIMO Performance Benefits

2x2 MIMO scheme doubles peak rate from 14.4 Mbps to 28.8 Mbps 2x2 MIMO provides significant experienced peak, mean & cell edge user data

rate benefits for isolated cells or noise/coverage limited cells 2x2 MIMO provides 20% – 60% larger spectral efficiency than 1x2

00.250.5

0.751

1.251.5

1.752

Near Cell Center Average CellLocation

Cell Edge

Dat

a R

ate

Gai

n o

f M

IMO

vs.

S

ISO

fo

ran

Iso

late

d C

ell SISO (1x1)

MIMO (2x2)Note: All gains normalized to Near Cell Center SISO Data Rate

0

20

40

60

80

100

Interference LimtedSystem

Isolated Cell

Spec

tral E

ffici

ency

Gai

n (%

) of 2

x2

MIM

O o

ver 1

x2 L

MM

SE


Overview of Dual Cell Operation

3GPP Rel.8 scope: The dual cell operation only applies to downlink HS-DSCH

Uplink traffic is carried on one frequency

The two cells belong to the same Node-B and are on adjacent carriers The two cells operate with a single TX antenna

Max two streams per user

Improvements in Rel.9 Dual-Band HSDPA MIMO in dual cell operation Dual Cell uplink

Multi-carrier HSDPA

Node-B

F1 UE F2

DL DL

5 MHz 5 MHz

UL 5 MHz

UL UTRAN configures one of the cell as the

serving cell for the uplink


Dual Cell HSDPA Operation for Load Balancing

Dual Cell HSDPA can optimally balance the load on two HSDPA carriers by scheduling active users simultaneously or on least loaded carrier at given TTI

Simple traffic and capacity model Avg. Transfer size:

1000 kbyte

Avg. Time between transfers: 60 sec

No gain at very high load

Dual Cell HSDPA operation versus Two legacy HSDPA carriers

0

1000

2000

3000

4000

5000

6000

7000

8000

0 10 20 30 40 50 60

Nb of users in sector footprint

Thro

ughp

ut in

kbp

s

Avg user throughput (2 HSDPA carriers)Avg Sector throughput (2 HSDPA carriers)Avg user throughput (Dual Cell HSDPA operation)Avg Sector throughput (Dual Cell HSDPA operation)


Enhanced Layer-2 Support for High Data Rates

Release 6 RLC layer cannot support

new peak rates offered by HSPA+ features such as MIMO & 64QAM RLC-AM peak rate limited to

~13 Mbps, even with aggressive settings for the RLC PDU size and RLC-AM window size

Release 7 introduces new Layer-2

features to improve HSDPA Flexible RLC PDU size MAC-ehs layer segmentation/

reassembly (based on radio conditions)

MAC-ehs layer flow multiplexing

Release 8 improves E-DCH MAC-i/ MAC-is

Rel.7

2

1500 byte IP packet

RLC-AM

MAC-ehs

RLC-AM PDU

1500

1502 3

MAC-ehs PDU

Traffic flow i for user k

2

1500 byte IP packet

RLC-AM

RLC-AM PDU

1500

1502

Traffic flow j for user k

3

22 bits

80 2 80 2

1500 byte IP packet

RLC-AM

80 2

MAC-hs

80 2 80 2 80 2

RLC-AM PDU

MAC-hs PDU

x19

..

.. Rel.6

Traffic flow i for user k


MAC-ehs in NodeB

MAC-ehs Functions Flow Control Scheduling/ Priority handling HARQ handling TFRC Selection Priority Queue Mux Segmentation

MAC-ehs

MAC – Control

HS-DSCH

TFRC selection

Priority Queue distribution

Associated Downlink Signalling

Associated Uplink Signalling

MAC-d flows

Priority Queue

Scheduling/Priority handling

Priority Queue

Priority Queue

Segmentation

Segmentation

Segmentation

Priority Queue MUX

HARQ entity

Cf. 25.321


HSDPA – UE Physical Layer Capabilities

HS-DSCH Category

Maximum number of HS-DSCH

multi-codes

Supported Modulation Formats

Minimum inter-TTI interval

Maximum MAC-hs TB size

Total number of soft channel

bits

Theoretical maximum data rate (Mbit/s)

Category 6 5 QPSK, 16QAM 1 7298 67200 3.6 Category 8 10 QPSK, 16QAM 1 14411 134400 7.2 Category 9 15 QPSK, 16QAM 1 20251 172800 10.1 Category 10 15 QPSK, 16QAM 1 27952 172800 14.0 Category 13 15 QPSK, 16QAM, 64QAM 1 35280 259200 17.6 Category 14 15 QPSK, 16QAM, 64QAM 1 42192 259200 21.1 Category 15 15 QPSK, 16QAM 1 23370 345600 23.3 Category 16 15 QPSK, 16QAM 1 27952 345600 28.0 Category 17 15 QPSK, 16QAM, 64QAM/

MIMO: QPSK, 16QAM 1 35280/

23370

259200/

345600

17.6/

23.3 Category 18 15 QPSK, 16QAM, 64QAM/

MIMO: QPSK, 16QAM 1 42192/

27952

259200/

345600

21.1/

28.0 Category 19 15 QPSK, 16QAM, 64QAM 1 35280 518400 35.2 Category 20 15 QPSK, 16QAM, 64QAM 1 42192 518400 42.2

Note: UEs of Categories 15 – 20 support MIMO cf. TS 25.306


E-DCH – UE Physical Layer Capabilities

E-DCH Category

Max. num. Codes

Min SF EDCH TTI Maximum MAC-e TB size

Theoretical maximum PHY data rate (Mbit/s)



14484/ 2798

1.45/ 1.4



20000/ 5772

2.0/ 2.89



20000/ 11484

2.0/ 5.74

Category 7 (Rel.7)

4 SF2 10 msec/ 2 msec

20000/ 22996

2.0/ 11.5

NOTE 1: When 4 codes are transmitted in parallel, two codes shall be transmitted with SF2 and two codes with SF4 NOTE 2: UE Category 7 supports 16QAM

cf. TS 25.306


Continuous Packet Connectivity (CPC)

Uplink DPCCH gating during inactivity significant reduction in UL interference

F-DPCH gating during inactivity

UE listens on HS-SCCH only when active

data control

HS-SCCH-less transmission introduced to reduce signaling bottleneck for real-time-services on HSDPA

Prior to Rel.7

Rel.7 using CPC

data control


CPC Performance Benefits

CPC provides up to a factor of two VoIP on HSPA capacity benefit

compared to R.99 AMR12.2 circuit voice and 35 – 40% benefit compared to Rel.6 VoIP on HSPA

0

0.5

1

1.5

2

2.5

3

AMR12.2 AMR7.95 AMR5.9VoIP

Cap

acity

Gai

n of

CPC

R'99 Circuit VoiceVoIP on HSPA (Rel'6)*VoIP on HSPA (CPC)*

Note: All capacity gains normalized to AMR12.2 Circuit Voice Capacity

* All VoIP on HSPA capacities assume two receive antennas in the terminal


“Always On” Enhancement of CPC

CPC allows UEs in CELL_DCH to “sleep” during periods of inactivity Reduces signaling load and battery consumption (in combination with DRX)

Allows users to be kept in CELL_DCH with HSPA bearers configured Need to page and re-establish bearers leads to call set up delay

Incoming request Page UE

Paging Response

Re-establish bearers

Send data

Without CPC, users typically kept in URA_PCH or CELL_PCH state to save radio resources and battery

UE in URA_PCH

CPC allows users to stay in CELL_DCH

Send data almost Immediately

(<50 ms reactivation)

Incoming request

UE in CELL_DCH

Avoids several hundred msec of call setup delay

CELL_FACH

CELL_DCH


Enhanced CELL_FACH & Enhanced Paging Procedure

UEs are not always kept in

CELL_DCH state, eventually fall back to URA_PCH/ CELL_PCH

HSPA+ introduces enhancements to reduce the delay in signaling the transition to CELL_DCH use of HSDPA in CELL_FACH and URA_PCH/ CELL_PCH states instead of S-CCPCH Enhanced CELL_FACH Enhanced Paging procedure

In Rel.8 improved RACH procedure

Direct use of HSUPA in CELL_FACH

Incoming request Page UE

Paging Response

Re-establish bearers

Send data

UE in URA_PCH

CELL_FACH

CELL_DCH

Use HSDPA for faster

transmission of signaling messages

2 msec frame length with up to 4 retransmissions


E-RACH – High level description

RACH preamble ramping as in R.99 with AICH/E-AICH acknowledgement Transition to E-DCH transmission in CELL_FACH

Possibility to seamlessly transfer to Cell_DCH NodeB can control common E-DCH resource in CELL_FACH

Resource assignment indicated from NodeB to UE

10 ms

#0 #1 #2 #3 #14#13#12#11#10#9#8#7#6#5#4#0 #1 #2 #3 #14#13#12#11#10#9#8#7#6#5#4τp-

a

#0 #1 #2 #3 #14#13#12#11#10#9#8#7#6#5#4#0 #1 #2 #3 #14#13#12#11#10#9#8#7#6#5#4PRACH access slots

10 ms

Access slot set 1 Access slot set 2

#0 #1 #2 #3 #14#13#12#11#10#9#8#7#6#5#4#0 #1 #2 #3 #14#13#12#11#10#9#8#7#6#5#4

Transmission starts with power ramping

on preamble reserved for E-DCH

access

NodeB responds by allocating common E-DCH

resources

UE starts common E-DCH transmission.

F-DPCH for power control, E-AGCH for rate control, E-HICH for HARQ


HSPA+ Architecture Evolution

Integration of some or all RNC functions into the NodeB provides benefits in terms of: Network simplicity (fewer network elements) Latency (fewer handshakes, particularly in combination with One-Tunnel) Synergy with LTE (serving GW, MME, eNB)

Backwards compatible with legacy terminals Central management of common resources

Control Plane

NodeB

RNC

SGSN

GGSN

Traditional HSPA Architecture

User Plane

NodeB

RNC

SGSN

GGSN

HSPA with One-Tunnel Architecture

NodeB+

SGSN

GGSN

HSPA+ with One-Tunnel Architecture for PS services


Evolved HSPA Architecture – Full RNC/NodeB collapse

2 deployment scenarios: standalone UTRAN or carrier sharing with “legacy”

UTRAN

Evolved HSPA NodeB

SGSN

GGSN

User plane : Iu / Gn ( ” one tunnel ” ) Control

plane : Iu

Iur

Evolved HSPA NodeB

SGSN

GGSN


plane : Iu

Iur

RNC

NodeB NodeB

” Legacy ” UTRAN

Iu

Evolved HSPA - stand - alone Evolved HSPA - with carrier sharing

NodeB+


plane : Iu

Iur

NodeB+


plane : Iu

RNC

NodeB NodeB

” Legacy ” UTRAN

Iu

Evolved HSPA - stand - alone Evolved HSPA - with carrier sharing


Home NodeB – Background

Home NodeB (aka Femtocell) located at the customers premise Connected via customers fixed

line (e.g. DSL) Small power (~100 mW) to

only provide coverage inside/ close to the building

Advantages Improved coverage esp. indoor Single device for home/ on the

move Special billing plans (e.g. home

zone) Challenges

Interference Security Costs

IP Network

UE

Gateway

Operator CN


CN Interface

Iuh

Iu-CS/PS

RAN Gateway Approach with new “Iuh” Interface

Approach Leverage Standard CN Interfaces

(Iu-CS/PS) Minimise functionality within

Gateway Move RNC Radio Control Functions

to Home NodeB and extend Iu NAS & RAN control layers over IP network

Features Security architecture Plug-and-Play approach Femto local control protocol CS User Plane protocol PS User Plane protocol HMS interface

RNC HNB-GW

NodeB HNB

Mobile CS/PS Core


HSPA References

Literature: H. Holma/ A. Toskala: “WCDMA for UMTS,” chapters 12 – 14, Wiley 2010 H. Holma/ A. Toskala (Ed.): “HSDPA/ HSUPA for UMTS,” Wiley 2006 Rhode und Schwartz: “Introduction to MIMO,” Application Note, July 2009 4G Americas: “The Evolution of HSPA,” White Paper, October 2011 H. Holma. A. Toskala, P. Tapia (Ed.): “HSPA+ Evolution to Release 12:

Performance and Optimization,” Wiley 2014 Standards

TS 25.xxx series: RAN Aspects TR 25.858 “HSDPA PHY Aspects” TR 25.308 “HSDPA: UTRAN Overall Description (Stage 2)” TR 25.877 “Iub/Iur protocol aspects” TR 25.896 “Feasibility Study for Enhanced Uplink for UTRA FDD” TR 25.808 “FDD Enhanced Uplink; Physical Layer Aspects” TR 25.319 “Enhanced Uplink: Overall Description (Stage 2)” TR 25.999 “HSPA Evolution beyond Release 7 (FDD)” TR 25.876 “Multiple-Input Multiple Output Antenna Processing for HSDPA” TR 25.903 “Continuous Connectivity for Packet Data Users” TR 25.820 “3G Home NodeB Study Item Technical Report”


Abbreviations

ACK (positive) Acknowledgement AG Absolute Grant AM Acknowledged (RLC) Mode AMC Adaptive Modulation & Coding BPSK Binary Phase Shift Keying BO Buffer Occupancy CAC Call Admission Control CPC Continuous Packet Connectivity CQI Channel Quality Indicator DC Dual Channel DDI Data Description Indicator FEC Forward Error Correction FIFO First In First Out FP Framing Protocol GoS Grade of Service HARQ Hybrid Automatic Repeat Request HOM Higher Order Modulation ID Identifier MAC Medium Access Control MIMO Multiple-Input Multiple-Output Mux Multiplexing NACK Negative Acknowledgement NBAP NodeB Application Part PCI Precoding Control Information PDU Protocol Data Unit PHY Physical Layer

PO Power Offset QoS Quality of Service QPSK Quadrature Phase Shift Keying RB Radio Bearer RG Relative Grant

RL Radio Link

RLC Radio Link Control RRC Radio Resource Control RRM Radio Resource Management RTT Round Trip Time SDU Service Data Unit SF Spreading Factor SG Serving Grant SI Scheduling Information SISO Single-Input Single-Output SM Spatial Multiplexing TB Transport Block TFC Transport Format Combination TFRC Transport Format & Resource Combination TFRI TFRC Indicator TPR Traffic to Pilot Ratio TTI Transmission Time Interval UM Unacknowledged (RLC) Mode 16QAM 16 (state) Quadrature Amplitude Modulation 64QAM 64 (state) Quadrature Amplitude Modulation VoIP Voice over Internet Protocol

Documents

High-Speed Downlink Packet Access (HSDPA) · Cellular Communication Systems Andreas Mitschele-Thiel, Jens Mueckenheim Nov. 2018 2 HSDPA Background Initial goals Establish a more spectral