Concepts of HSDPA

Abstract: The W-CDMA specifications are the most ambitious and complex radio standards ever written. The addition of new features to improve the system thus comes as no surprise. One of the most significant changes since Release 99 is the incorporation of High Speed Downlink Packet Access (HSDPA). This technology allows for high data rates in the downlink, which enables the WCDMA system to compete against other high data rate wireless technologies. This presentation covers the main changes that HSDPA introduces in the physical and transport layers, such as the structure of the new physical and transport channels and the new link adaptation techniques. This paper is also available as an Agilent Application Note: Concepts of High Speed Downlink Packet Access: Bringing Increased Throughput and Efficiency to W-CDMA (literature part number 5989-2365EN)

3GPP Specifications Releases

GSM Phase I Original GSM GSM Phase II Doc. Version 4.X.Y DCS1800 Release 96 HSCSD Release 97 GPRS Release 98 EDGE
Concepts of HSDPA rev.8.1 November 9th, 2005 Agilent Copyright Page 2

Doc. Version 3.X.Y

Release 99 Release 4

Doc. Version 3.X.Y Doc. Version 4.X.Y

W-CDMA (FDD and wideband TDD) Narrowband TDD (Low Chip Rate TD-SCDMA) Release 5 Doc. Version 5.X.Y High Speed Downlink Packet Access (HSDPA) Release 6 Doc. Version 7.X.Y Release 7 Doc. Version 6.X.Y Doc. Version 7.X.Y

Doc. Version 5.X.Y Doc. Version 6.X.Y

In December 1999, ETSI produced a CD releasing the Frozen version of W-CDMA R99. During 2000, the R99 feature set was frozen with regard to new features, but since Dec 99 there have been over 6500 non-editorial changes. The scale of change and the fact that it is still going on means there are now strict rules to limit changes to the absolute essentials. The effect of this policy is that issues regarding omissions or ambiguity with the R99 standard are now being left out or deferred to future releases like Release 5, 6 and beyond. The release naming of the ETSI/3GPP standards is shown below. Major features for each release are also noted: GSM Phase 1 Doc. Version 3.X.Y GSM Phase II Doc. Version 4.X.Y Release 96 Release 97 Release 98 Release 99 Release 4 Release 5 Release 6 Release 7 Doc. Version 5.X.Y Doc. Version 6.X.Y Doc. Version 7.X.Y Doc Version 3.X.Y Doc. Version 4.X.Y Doc. Version 5.X.Y Doc. Version 6.X.Y Doc. Version 7.X.Y Original GSM DCS1800 HSCSD GPRS EDGE W-CDMA - Content frozen in Dec. 99 Narrowband TDD Content frozen in March 01 HSDPA Content frozen in June 02 Content to be frozen in the near future Content and dates to be defined

After R99, all the GSM documents were renamed by adding 40 to the document number to avoid confusion between GSM Phase II with Release 4, and Release 96 with release 5 etc. This presentation focuses on HSDPA technology as defined in Release 5.

What is HSDPA?
HSDPA : High Speed Downlink Packet Access Increases downlink data rates
Theoretically up to a peak data rate of almost 14 Mbps In practice, peak data rate might be 3.6 Mbps or lower, depending on cell size

Improves spectral efficiency for downlink packet data services Employs both Time Domain Multiplexing (TDM) and Code Domain Multiplexing (CDM) Reduces transmission latency Supports multimedia services (packet based): web browsing, movie streaming, interactive games, car-navigation system update, etc HSDPA User Equipment (UE) and Release 99 UE can share the same carrier
Concepts of HSDPA rev.8.1 November 9th, 2005 Agilent Copyright Page 3

W-CDMA technology, can in perfect conditions deliver peak data rates of up to 2 Mbps. But in typical network deployment, a cell will have a maximum capacity of around 1 Mbps shared between the cells users. Peak user data rates are limited to 384 kbps. HSDPA offers peak theoretical downlink data rates of up to 14 Mbps and increases the system capacity for downlink packet data. Time Domain Multiplexing (TDM) is used as well as Code Domain Multiplexing (CDM), which accommodates bursty packet data better and makes more efficient use of the spectrum. Although the theoretical peak data rate for HSDPA is approximately 14 Mbps, the actual rates achieved will be much lower than that. The performance of HSDPA depends largely on the cell size. In macro cell applications, HSDPA may improve on W-CDMA data capacity only by perhaps 30%, with sustainable peak data rates for one user of maybe 1 Mbps. But in micro and pico cell deployments where co-channel interference is minimal, HSDPA is capable of delivering much higher performance over basic W-CDMA. The exact improvement is very hard to predict since it depends on actual channel conditions and the real-time capabilities of the BTS neither of which are standardized. However, some credible estimates for Release 5 suggest a cell capacity of up to 3 Mbps rising to 5 Mbps in Release 6, which includes a more advanced UE receiver and improved BTS packet scheduling. Peak user data rates might reach 3.6 Mbps for short periods of time but are unlikely to be sustainable. The increased data rates and improved capacity result in shorter delays for the end-users. This is particularly important for some multimedia applications such as interactive games. The high data rates also benefit streaming and web browsing applications. Because HSDPA technology is backwards compatible with 3GPP Release 99, voice and data applications developed for W-CDMA can still be run on the upgraded networks, and the same radio channel will support W-CDMA and HSDPA services simultaneously.

Shorter Radio Frame Downlink Channels

Whats New for HSDPA?

16 QAM Modulation Type UEs Share Code Space by Time Multiplexing Uplink Channel Adaptive Modulation and Coding (AMC) Hybrid-ARQ (HARQ) Medium Access Control (MAC) Scheduling Functionality Moved to Node-B Test Model 5 and Fixed Measurement Channels What is next? - EUDCH and MIMO Agilent Solutions for HSDPA Testing
Concepts of HSDPA rev.8.1 November 9th, 2005 Agilent Copyright Page 4

This slide shows the main changes that HSDPA introduces. Each of these areas will be presented in greater detail. The presentation focuses on the physical layer aspects of HSDPA (for the FDD mode), so the Medium Access Control (MAC) functionality will not be covered in depth.

Sub-Frame Structure (FDD)

Sub-frame of Length 2 ms (3 slots, 3 x 666.67us)
W-CDMA frame: 10 ms
Subframe #0 Subframe #1 Subframe #2 Subframe #3 Subframe #4

HSDPA subframe: 3 x Tslot = 2ms, 7680 chips

Concepts of HSDPA rev.8.1 November 9th, 2005

Agilent Copyright

Page 5

The HSDPA frame (sub-frame) is 2 ms in length. That is equivalent to 3 of the currently defined W-CDMA slots. Therefore there are 5 HSDPA sub-frames per a 10 ms WCDMA frame.

Channel Types
New Physical Channels: DOWNLINK
HS-PDSCH : High Speed Physical Downlink Shared Channel Carries payload data HS-SCCH : High Speed Shared Control Channel Carries UE identity and channel parameters of associated HS-PDSCH

UPLINK
HS-DPCCH : High Speed Dedicated Physical Control Channel Carries ACK and CQI (Channel Quality Indicator)

New Transport Channel (DOWNLINK only):

HS-DSCH : High Speed Downlink Shared Channel
Concepts of HSDPA rev.8.1 November 9th, 2005 Agilent Copyright Page 6

For HSDPA, one downlink transport channel (High Speed Downlink Shared Channel HS-DSCH), two downlink physical channels (High Speed Physical Downlink Shared Channel HS-PDSCH and High Speed Shared Control Channel HS-SCCH), and one uplink physical channel (High Speed Dedicated Physical Control Channel HS-DPCCH) have been introduced. With this enhancement, layer 2 (MAC layer) can map logical channels (DCCH and DTCH) onto the transport channel (HS-DSCH). Then, layer 1 in turn maps the transport channel (HS-DSCH) onto one or more physical channels (HS-PDSCH). The physical layer creates HS-SCCH and HS-DPCCH to control and assist HS-DSCH transmission.

Downlink Physical Channel Structure HS-PDSCH

Sub-frame Length 2 ms Two Slot Formats (based on modulation type)
Data T slot = 2560 chips, M*160 bits

Slot #0

Slot #1 1 HS-PDSCH subframe: Tf = 2 ms

Slot #2

M is the number of bits per modulation symbol i.e. M=2 for QPSK and M=4 for 16QAM.
Slot format #i Channel Bit Rate (kbps) Channel Symbol Rate (ksps) SF Bits/ HS-DSCH subframe Bits/ Slot 0(QPSK) 480 240 16 960 320 1(16QAM) 960 240 16 1920 640

Ref : 25.211 5.3.3.13

Concepts of HSDPA rev.8.1 November 9th, 2005 Agilent Copyright Page 7

The HS-PDSCH is used to carry the HS-DSCH. A HS-PDSCH is assigned one channelization code of fixed spreading factor SF=16 from the set of channelization codes reserved for HS-DSCH transmission. Multi-code transmission is allowed, which translates to the UE being assigned multiple channelization codes (multiple HS-PDSCHs) in the same HS-PDSCH subframe, depending on its UE capability. The symbol rate of HS-PDSCH is fixed (240ksps), but its code number is configurable. An HS-PDSCH may use QPSK or 16QAM modulation, so its channel bit rate can vary (480 kbps vs. 960 kbps) based on its modulation scheme (QPSK vs. 16QAM). The channel bit rate corresponds to the data rate after coding is applied. Originally 8 PSK and 64 QAM were investigated, but are not part of the Release 5 specifications. All relevant Layer 1 information is transmitted in the associated HS-SCCH, i.e. the HSPDSCH does not carry any Layer 1 information.

16QAM for HS-PDSCH

downlink physical channel

S P

I
Modulation Mapper

Sdl,n
I+jQ

i1

Cch,SF,m
Q

S
q2

i2
1011 1001
1.3416

i2
0001 0011

1010

1000
0.4472

0000
0.4472

0010
1.3416

1110

1100

0100

0110

q1 q2
1111 1101 0101 0111

Left (16QAM): 240ksps, 960 kbps Right (QPSK) : 240ksps, 480 kbps
Ref : 25.213 5.1
Concepts of HSDPA rev.8.1 November 9th, 2005 Agilent Copyright Page 8

The HS-PDSCH is modulated, spread, scrambled, and summed like other WCDMA downlink physical channels. The only difference is that the modulation mapping can be either for QPSK or for 16QAM, and that the SF is always fixed at 16. The top right IQ constellation shows the correspondence of symbols to bits, as described in the specifications. The code domain displays for two signals with 4 HS-PDSCHs are shown in the slide. The displays also show the constellation for one of the HS-PDSCHs, which has a different modulation scheme for each signal: 16QAM modulation (left) and QPSK modulation (right).

Downlink Transport Channel Coding HS-DSCH

Number of Transport Blocks: Max = 1 CRC Length = 24 bits Channel Coding = 1/3 Rate Turbo Coding Hybrid-ARQ is Two Stage Rate Matching Interleaving is Performed Separately For Each Physical Channel Different 16QAM Constellations are Defined
Ref : 25.212 4.5
Concepts of HSDPA rev.8.1 November 9th, 2005 Agilent Copyright

CRC attachment Bit Scrambling Code block segmentation Channel Coding

Phys. Layer Hybrid-ARQ Functionality Physical channel segmentation HS-DSCH Interleaving Constellation Re-arrangement (16QAM) Physical channel mapping PhCH#1 PhCH#P
Page 9

The HS-DSCH is an evolution of the downlink shared channel (DSCH) in W-CDMA Release 99. As opposed to the dedicated channel (DCH), the DSCH provides the possibility of time-multiplexing different users. In order to obtain higher data rates and improve spectral efficiency, on the HS-DSCH the fast power control and variable spreading factor of the DSCH are replaced by short packet size, AMC, multi-code operation, and HARQ, as will be shown during this presentation. The HS-DSCH exists only in the downlink, and is always associated with a DPCH and one or more shared physical control channels (HS-SCCHs). There is only one HS-DSCH per Coded Composite Transport Channel (CCTrCH). There is only one CCTrCH of HS-DSCH type per UE. The CCTrCH can be mapped onto one or several physical channels (HS-PDSCHs). The coding blocks for the HS-DSCH are shown above. The channel coding of the HS-DSCH is always 1/3 (based on the Release 99 1/3 Turbo encoder). The effective code rate varies though, depending on the puncturing and repetition that is applied during the rate matching stage. The HARQ functionality corresponds to a two-stage rate matching. It matches number of bits at the output of the channel coder to the total number of bits of the HS-PDSCH set to which the HS-DSCH is mapped. The HARQ functionality is controlled by the redundancy version (RV) parameters. The exact set of bits at the output depends on the number of input bits, the number of output bits, and the RV parameters. Physical channel segmentation divides the bits among the different physical channels, when more than one HS-PDSCH is used. Interleaving is done separately for each physical channel. The constellation re-arrangement only applies to 16QAM, so it is transparent for QPSK. In 16QAM modulation, two of the four bits in a symbol have a higher probability of error than the two other bits. The re-arrangement is applied during retransmissions and provides equal probability of error to all the bits in average, after the retransmission combining.

Channel Coding Example FRC H-Set 4 HS-DSCH

Inf. Bit Payload CRC Addition Code Block Segmentation Turbo-Encoding (R=1/3) 1st Rate Matching RV Selection Physical Channel Segmentation
4800 7200 3202 3202 24 CRC

Transport Block Size : 3202 bits

3226

Effective code rate : 3226/4800 = 0.67

9678 1 2 Tail Bits

1st rate matching

2nd rate matching

960

Physical Channel Bits/TTI : 4800 bits Virtual IR buffer size : 7200 bits/process
Ref : 25.101 A.7.1.4
Page 10

5 HS-PDSCHs QPSK mod. (960 bits/TTI)

Concepts of HSDPA rev.8.1 November 9th, 2005 Agilent Copyright

Spreading Scrambling

An example of the coding for a HS-DSCH is shown above. The coding corresponds to a Fixed Reference Channel (FRC H-Set 4), that is used for UE receiver testing, as will be shown later. The first rate matching stage matches the number of input bits to a virtual incremental redundancy (IR) buffer. The second rate matching stage matches the number of bits after the first matching stage to the number of physical channel bits available in the HS-PDSCH set in the Transmission Time Interval (TTI) and it is controlled by the RV parameters. The number of HS-PDSCH channels (five in this example) and the modulation format define the number of physical channel bits after RV selection (960 bits x 5 = 4800 bits) The turbo-encoding code rate is fixed at 1/3. The effective code rate corresponds to the combination of turbo-encoding, and the rate matching stages. So, the effective code rate for any HS-DSCH configuration can be calculated if the transport block size (which in this case is equivalent to the number of information bits), the number of HS-PDSCHs, and the modulation format are known. In this case the effective code rate is 0.67 (= (3202+24) bits / (960 bits x 5)).

10

Downlink Physical Channel Structure HS-SCCH

Carries Downlink Signalling Related to HS-DSCH Transmission UE Monitors up to 4 HS-SCCHs Simultaneously (HS-SCCH set)
Control Data T slot = 2560 chips, 40 bits

HS-SCCH
Slot #2

Slot #0

Slot#1 1 HS-SCCH subframe: Tf = 2 ms

Modulation QPSK
Concepts of HSDPA rev.8.1 November 9th, 2005

Channel Bit Rate (kbps) 60

Agilent Copyright

Channel Symbol Rate (ksps) 30

SF 128

Bits/ Slot 40

Ref : 25.211 5.3.3.12

Page 11

The HS-SCCH carries signaling information related to the HS-DSCH transport channel transmission. The bit rate of HS-SCCH is fixed at 60 kbps, but its code number is configurable. A base station (BTS) can transmit more than 4 HS-SCCHs, but any single UE is required to monitor 4 (see 25.308 5.2.2). There is a maximum of 32 HS-SCCHs per cell. (see 25.433 maxNrOfHSSCCHs) The HS-SCCH is also spread and summed like other downlink physical channels.

11

Downlink Physical Channel Structure Control Data HS-SCCH

T slot = 2560 chips, 40 bits

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Cch,16,0 1 1 1 1 1 1 1 1-1-1-1-1-1-1-1-1 Cch,16,1 Slot #0 Slot#1 Slot #2 1 1 1 1-1-1-1-1 1 1 1 1-1-1-1-1 Cch,16,2 O(Offset) 1 1 1 1-1-1-1-1-1-1-1-1 1 1 1 1 Cch,16,3 1 1-1-1 1 1-1-1 1 1-1-1 1 1-1-1 Cch,16,4 P(# of codes) 1 HS-SCCH subframe: Tf = 2 ms 1 1-1-1 1 1-1-1-1-1 1 1-1-1 1 1 Cch,16,5 1 1-1-1-1-1 1 1 1 1-1-1-1-1 1 1 Cch,16,6 HS-PDSCH 1 1-1-1-1-1 1 1-1-1 1 1 1 1-1-1 Cch,16,7 Code Slot #0 carries modulation information of HS-PDSCH, such as: 1-1 1-1 1-1 1-1 1-1 1-1 1-1 1-1 Cch,16,8 1-1 1-1 1-1 1-1-1 1-1 1-1 1-1 1 Cch,16,9 Set - Channelization Code Set (7 bits) 1-1 1-1-1 1-1 1 1-1 1-1-1 1-1 1 Cch,16,10 - Modulation Scheme: QPSK or 16QAM (1 bit) 1-1 1-1-1 1-1 1-1 1-1 1 1-1 1-1 Cch,16,11 1-1-1 1 1-1-1 1 1-1-1 1 1-1-1 1 Cch,16,12 1-1-1 1 1-1-1 1-1 1 1-1-1 1 1-1 Cch,16,13 1-1-1 1-1 1 1-1 1-1-1 1-1 1 1-1 Cch,16,14 Slot#1,Slot#2 carries channel-coding information of HS-DSCH, such as: 1-1-1 1-1 1 1-1-1 1 1-1 1-1-1 1 Cch,16,15

- Transport-block size information (6 bits) - Hybrid-ARQ process information (3 bits) - Redundancy and constellation version (3 bits) - New data indicator (1 bit)

Ref : 25.212 4.6

Concepts of HSDPA rev.8.1 November 9th, 2005 Agilent Copyright

Data of Slot #0,#1,#2 are covered with UE identity (16 bits)

Page 12

The physical channel structure of the HS-SCCH is shown above. The first slot carries critical information for HS-PDSCH reception, such as the channelization code set and the modulation scheme. After receiving the first slot, the UE has just one time slot for decoding the information and preparing for the upcoming HS-PDSCH reception. The number of codes used for HSDPA can vary dynamically between 1 and 15. The Orthogonal Variable Spreading Factor (OVSF) codes are positioned in sequence in the code tree. The number of multicodes and the corresponding offset for the HS-PDSCHs mapped from a given HS-DSCH is signaled on the HSSCCH. P Multicodes at Offset O are Allocated as Follows: Cch,16,O Cch,16,O+P-1. The second and third slots carry channel coding information of the HS-DSCH, such as the transport block size information, the HARQ information, the RV and constellation version, and the new data indicator. The data of the three slots are covered with the UE identity.

12

Downlink Physical Channel Timing HS-SCCH/HS-PDSCH

HS-PDSCH Starts 2 Slots (5120 chips) After the Start of the HS-SCCH Start of HS-SCCH Sub-frame #0 is Aligned with the Start of the P-CCPCH Frames
3Tslot = 7680 chips HS-SCCH 3Tslot 7680 chips HS-PDSCH HS-DSCH sub-frame HS-PDSCH (2Tslot = 5120 chips)
1 sub frame = 3 slots (=1/5 frame) duration : 2 ms (=10 ms * 1/5)

Ref : 25.211 7.8

Page 13

Concepts of HSDPA rev.8.1 November 9th, 2005

Agilent Copyright

This timing chart above shows how control information associated with a HSPDSCH subframe is provided on a HS-SCCH subframe by BTS. BTS transmits a HS-SCCH subframe two slots (= 5,120 chips) before transmitting the HS-PDSCH. Therefore, the UE needs to decode the HS-SCCH subframe quickly in a short limited time. The HS-SCCH and the HS-PDSCH overlap during a slot. Therefore, indispensable information for demodulating the HS-PDSCH subframe (e.g. modulation scheme) needs to be transmitted at the beginning of the HS-SCCH subframe. Otherwise, the UE would miss the HS-PDSCH.

13

Downlink CODE/TIME multiplexing

Shared packet channel Time Dedicated channel OVSF codes

Concepts of HSDPA rev.8.1 November 9th, 2005 Agilent Copyright Page 14

This picture shows how both Time Domain Multiplexing (TDM) and Code Domain Multiplexing (CDM) can be used in HSDPA. Multiple UEs can receive information in the same sub-frame by using different parts of the code (OVSF) space. The main difference between a dedicated channel (such as the DPCH used in WCDMA Release 99) and a shared packet channel (such as the HS-PDSCH used in HSDPA) is that shared packet channels can also multiplex several users in the time domain.

14

Uplink Physical Channel Structure HS-DPCCH

Carries Feedback Related to Downlink HS-DSCH Transmission Can Only Exist with DPCCH
Tslot = 2560 chips HARQ-ACK 2 x Tslot = 5120 chips CQI

1 HS-DPCCH subframe (2 ms)

Subframe #0 Subframe #i Subframe #4

1 Radio frame Tf = 10 ms
Slot Format #i 0
Concepts of HSDPA rev.8.1 November 9th, 2005

Ref : 25.211 5.2.1

Transmitted slots per Subframe 3
Page 15

Channel Bit Rate (kbps) 15

Agilent Copyright

Channel Symbol SF Bits/ Bits/ Rate (ksps) Subframe Slot 15 256 30 10

The HS-DPCCH carries uplink feedback signaling related to downlink HS-DSCH transmission. The HS-DSCH-related feedback signaling consists of HARQ Acknowledgement (HARQ-ACK) and Channel-Quality Indication (CQI). As in the case of the HS-PDSCH and the HS-SCCH, each sub frame of length 2 ms (3*2560 chips) consists of 3 slots, each of length 2560 chips. The HARQ-ACK is carried in the first slot of the HS-DPCCH sub-frame. The CQI is carried in the second and third slot of a HSDPCCH sub-frame. There is at most one HS-DPCCH on each radio link. The HS-DPCCH can only exist together with an uplink DPCCH. Two different paths are used for coding: HARQ-ACK and CQI. The HARQ-ACK information (1 bit) is coded to 10 bits, being ACK coded as 1s, and NACK coded as 0s. The CQI information is coded Using a (20,5) Code. The coded bits are directly mapped to the physical channel. The feedback cycle of the CQI can be set as a network parameter in predefined steps from 2ms to infinite (i.e. disabled). It is worth noticing that an active HS-DPCCH may have slots where no HARQ-ACK information and/or CQI information is transmitted. Therefore, the HS-DPCCH can carry a bursted signal.

15

Uplink Spreading/Scrambling HS-DPCCH

DPDCH1 DPDCH3 DPDCH 5 HS-DPCCH* cd,1 cd,3 cd,5 chs d d d hs

* If Nmax-dpdch is even ** If Nmax-dpdch is odd

Sdpch,n I+jQ S

DPDCH 2 DPDCH 4 DPDCH 6 DPCCH HS-DPCCH**

cd,2 cd,4 cd,6 cc chs

d d d c hs
Agilent Copyright

Q
j

channelization code of HS-DPCCH Nmax-dpdch Channelization code Cch 1 2,4,6 3,5 Cch,256,64 Cch,256,1 Cch,256,32

HS-DPCCH (Cch,256,32) Ref : 25.213 4.2.1 25.213 4.3.1.2 25.214 5.1.2.6 for power
Page 16

Concepts of HSDPA rev.8.1 November 9th, 2005

The diagram for uplink spreading and scrambling is shown above. There are two possible HS-DPCCH positions on the summation block. If the number of DPDCH is even, HS-DPCCH goes to I-summation. If the number of DPDCH is odd, HS-DPCCH goes to Q-summation. In order to balance the total power on I and Q axes, the HS-DPCCHs channelization code varies depending on the DPDCH channels activity status. The table above shows that the channelization code of the HS-DPCCH depends on the number of DPDCHs being used. (e.g. if only 1 DPDCH is used, the HSDPCCH would be on the Q path with a channelization code of 64; if there are three DPDCHs, the HS-DPCCHs code number is 32, and since three is an odd number, HS-DPCCH goes to Q-axis.) The power level of the HS-DPCCH is set as an offset compared to the UL DPCCH. The bhs value is derived from the power offsets of the HARQ-ACK and CQI fields, which are signalled by the higher layers.

16

Radio Frame Timing of Downlink and Uplink Physical Channels

At BTS
10msec Radio frame Any CPICH P-CCPCH
` n:th DL DPCH

Radio frame (SFN mod 2 = 0)

Radio frame (SFN mod 2 = 1)

DPCH, n
0 0 1 1 2 2 3 3 4 4 0 0 1 1 2 2 3 3 4 4

Ref : 25.211 Figure33 Ref : 25.211 7.7 Ref : 25.211 7.8 Ref : 25.402 8

HS-SCCH subframe # HS-PDSCH subframe # Propagation delay

At UE

n:th DL DPCH UL DPCH

DPCH, n 1024 chip+/-alpha TTX_diff

0 1 2 3 0 4 1 0 2 1 3 2 4 3 4

DL HS-PDSCH subframe # UL HS-DPCCH subframe #

UEP 19200 chips = 7.5 slots

m = (TTX_diff / 256) + 101

m * 256 chips = 7680 + 19200 + T TX_diff - 1024

Concepts of HSDPA rev.8.1 November 9th, 2005 Agilent Copyright Page 17

The radio frame timing for the downlink and uplink physical channels is shown above. The HS-SCCH subframe #0 aligns with the W-CDMA 10ms radio frame. The HS-DPCCH is transmitted approximately 19,200 chips (= 7.5 slots) after its corresponding downlink HS-PDSCH subframe. The time of the ACK/NACK response from the UE on the HS-DPCCH is explicitly defined. The BTS will know when to expect the response for each sub-frame it transmitted. The HS-DPCCH transmission time is referenced to the UL DPCH and the transmit time difference between the DL DPCH and the sub-frame of interest. The overall time from the beginning of DL HS-PDSCH sub-frame until the beginning of the UL HS-DPCCH sub-frame is a little greater than 3 sub-frames. This timing information will be important later when discussing the number of HARQ processes.

17

Which of the following statements about dedicated and shared transport channels are correct?
Dedicated transport channels are code multiplexed with other channels. A dedicated transport channel always uses the same code(s) and shares it/them in time with other users. A shared transport channel always uses the same code(s) and shares it/them in time with other users. Shared transport channels share the time but not the code space with other users.
Agilent Copyright Page 18

Concepts of HSDPA rev.8.1 November 9th, 2005

Information about the Q&A (review slide 14 for further information):

Which of the following statements about dedicated and shared transport channels are correct? Dedicated transport channels are code multiplexed with other channels. True. This is the basic principle of CDMA technology. A dedicated transport channel always uses the same code(s) and shares it/them in time with other users. False. A dedicated channel does not share its code channel with other users in time. So, no TDM is used. A shared transport channel always uses the same code(s) and shares it/them in time with other users. False. The code set (including the number of codes) for a shared transport channel can change dynamically. Shared transport channels share the time but not the code space with other users. False. Shared transport channels share both the time and the code space with other users. So, both CDM and TDM are used.

18

Which of the following are true differences between the HS-DSCH and the DSCH?
The HS-DSCH has a shorter frame size than the DSCH The HS-DSCH uses TDM and CDM and the DSCH only CDM There is fast power control for the DSCH but not for the HSDSCH The HS-DSCH uses codes with a fixed SF and the DSCH uses codes with a variable SF

Concepts of HSDPA rev.8.1 November 9th, 2005

Agilent Copyright

Page 19

Information about the Q&A (review slide 9 for further information):

Which of the following are true differences between the HS-DSCH and the DSCH? The HS-DSCH has a shorter frame size than the DSCH True. The DSCH is a basic W-CDMA channel, so it has a 10 s frame. The HS-DSCH uses the 2 ms HSDPA subframe. The HS-DSCH uses TDM and CDM and the DSCH only CDM False. They are both shared channels, so they use both TDM and CDM. There is fast power control for the DSCH but not for the HS-DSCH True. The fast power control and variable spreading factor used in basic W-CDMA has been replaced in the HSDPA HS-DSCH channel with a shorter frame size, fixed spreading factor, and the use of techniques such as AMC and HARQ. The HS-DSCH uses codes with a fixed SF and the DSCH uses codes with a variable SF True. The HS-DSCH is always mapped onto physical channels (HS-PDSCH) with SF=16. The DSCH used in basic W-CDMA is mapped onto physical channels with variable spreading factors.

19

What is the maximum number of HS-PDSCHs for an HS-DSCH within a subframe?

1 3 15 16

Concepts of HSDPA rev.8.1 November 9th, 2005

Agilent Copyright

Page 20

Information about the Q&A (review slide 7 for further information):

What is the maximum number of HS-PDSCHs for an HS-DSCH within a subframe? 1 False. 3 False. 15 True. The HS-PDSCHs use SF=16, so if no other channels were present, there could be 16 of these channels. Since other channels, such as the HS-SCCH, and the regular WCDMA channels (CPICH, P-CCPCH, etc) need to be present and occupy some code space as well, there can be up to 15 HS-PDSCHs. 16 True.

20

HS-SCCH slot #0 carries

HARQ ACK/NACK info The CQI HS-PDSCH modulation information The HS-PDSCH redundancy and constellation version

Concepts of HSDPA rev.8.1 November 9th, 2005

Agilent Copyright

Page 21

Information about the Q&A (review slide 12 for further information):

HS-SCCH slot #0 carries HARQ ACK/NACK info False. The HARQ ACK/NACK info is carried by the uplink HS-DPCCH The CQI False. The CQI is carried by the uplink HS-DPCCH HS-PDSCH modulation information True. HS-SCCH slot 0 carries both information about the code set and the modulation used for the HS-PDSCH(s) onto which the HS-DSCH is mapped. The HS-PDSCH redundancy and constellation version. False. The HS-DSCH redundancy and constellation version, along with other transport channel coding information, is carried by slot #1 and slot #2 of the HS-SCCH.

21

Which of the following statements about the HSDPCCH are true?

The HS-DPCCH always uses SF=256 The HS-DPCCH is applied to I or Q path depending on the number of HS-PDSCHs transmitted The HS-DPCCH carries HARQ ACK/NACK information As in the other uplink channels, the HS-DPCCH modulation scheme is always BPSK.

Concepts of HSDPA rev.8.1 November 9th, 2005

Agilent Copyright

Page 22

Information about the Q&A (review slides 15 and 16 for further information):
Which of the following statements about the HS-DPCCH are true? The HS-DPCCH always uses SF=256 True. Both uplink control channels (the DPCCH and the HS-DPCCH) use always SF=256. The HS-DPCCH is applied to I or Q path depending on the number of HS-PDSCHs transmitted False. The HS-DPCCH is applied to either the I or Q path depending on the number of uplink DPDCHs (not downlink HS-PDSCHs!!) transmitted. The HS-DPCCH carries HARQ ACK/NACK information True. Slot #0 of the HS-DPCCH carries ACK/NACK info. As in the other uplink channels, the HS-DPCCH modulation scheme is always BPSK. True. All the uplink channels (basic W-CDMA or HSDPA) are I/Q multiplexed and use BPSK.

22

Adaptive Modulation and Coding

Change Modulation and Coding With Instantaneous Channel Conditions Channel Conditions Reported by CQI in HS-DPCCH Modulation and Coding Capable of Being Adjusted Every Sub-frame (2 ms)
Example: Coding parameters from CQI table for UE categories 1 to 6
CQI Value 0 1 15 16 29 30 137 3319 3565 7168 7168 1 5 5 5 5 Transport block size Number of HS-PDSCHs Modulation Reference power adjustment dB 0 0 0 -7 -8

Virtual IR buffer size

NIR RV

Out of range QPSK QPSK 16QAM 16QAM 16QAM 9600 9600 9600 9600 9600 0 0 0 0 0

Ref : 25.214 6.A.2 and Tables 7A-E

Concepts of HSDPA rev.8.1 November 9th, 2005 Agilent Copyright Page 23

Link adaptation is one important way in which HSDPA improves data throughput. The technique used, adaptive modulation and coding (AMC), varies the downlink modulation-coding scheme depending on the channel conditions for each user. The power of the transmitted signal is held constant over a sub-frame interval, and the modulation and coding format are changed to match the current received signal quality or channel conditions at the receiver. In this scenario, users close to the BTS are typically assigned higher order modulation with higher code rates (for example, 16QAM with an effective code rate of 0.89), but the modulation-order and code rate will decrease as the distance from the BTS increases. As mentioned earlier, 1/3 rate Turbo coding is used and different effective code rates are obtained through various rate-matching parameters. In HSDPA, the UE reports the channel conditions to the BTS via the uplink channel CQI field in the HSDPCCH. The CQI value can be 0 to 30, with a value of 0 indicating out-of-range. Each CQI value corresponds to a certain transport block size, number of HS-PDSCHs, modulation format, reference power adjustment ? , virtual IR buffer size and RV parameter, for a certain UE category(the different UE categories will be defined later in this presentation). The UE reports the maximum CQI value whose corresponding parameters would theoretically provide an acceptable block error ratio (BLER) for the current link conditions. The reported CQI value is then used by the BTS in combination with other parameters to determine the appropriate coding configuration for the next packet transmission to the UE. To fully define the coding configuration for a certain transmission, the BTS must select the following parameters: Transport block size (254 from which to choose in the FDD mode - see 25.321 Appendix A) Modulation typeQPSK or 16 QAM Number of physical downlink codes (1-15) Rate-matching parameters: Virtual IR buffer size and RV (or data puncturing scheme). So, even though there are a limited number of CQI values, there are literally thousands of configurations from which the BTS can choose for a certain transmission.

23

Effective Code Rate and Nominal Data Rate

Effective _ code _ rate =

(TBsize + 24)
(2)
Effective Code Rate 0.17 0.20 0.35 0.75 0.89 Nominal Data Rate 68.5 kbps 86.5 kbps 325 kbps 3.584 Mbps 12.779 Mbps

(# HS PDSCH # bitsHS PDSCH )

(1)

Nom. _ Data _ Rate =

For example:
Transport block size 137 173 650 7168 25558
Concepts of HSDPA rev.8.1 November 9th, 2005

TBsize 2ms

Number of HS-PDSCHs 1 1 2 5 15
Agilent Copyright

Modulation QPSK QPSK QPSK 16QAM 16QAM

Page 24

The transport block size, number of HS-PDSCHs and modulation scheme determine the effective code rate as shown in equation (1) in the slide. The nominal data rate can be directly calculated from the transport block size divided by the Transmission Time Interval (TTI), which in HSDPA is fixed at 2 ms, as shown in equation 2. Therefore, in HSDPA many data rates (or transport block sizes) are possible by varying the modulation type, effective code rate, and number of multicodes. This slide shows a few examples. The modulation, coding, and number of HS-PDSCHs can be adjusted every sub-frame (2 ms). Remember that even though there are only 30 CQI values, there are 254 possible data rates (transport block sizes) for the FDD mode. There are almost 2000 possible effective code rates (almost 2000 combinations of the three parameters: transport block size, number of HS-PDSCHs, and modulation scheme).

24

Maximum Data Rate

Transport Block Size 27952 Number of HS-PDSCHs 15 Modulation 16QAM Effective Code Rate 0.97 Nominal Data Rate 13.976 Mbps

15 HS-PDSCH

Concepts of HSDPA rev.8.1 November 9th, 2005

Agilent Copyright

Page 25

For example, the largest transport block size is 27952 bits. It corresponds to a highest data rate of 13.976 Mbps (27952 bits/2 ms = 13.976 Mbps). This data rate is obtained by using 16QAM, an effective code rate of 0.9714, and 15 HS-PDSCHs. In real life, the 14 Mbps headline figure for HSDPA is not achievable. It is physically possible to configure such a channel but there is nowhere it could be used. It requires close-to-perfect link conditions. The code domain display that shows the 15 HS-PDSCHs with the 16 QAM constellation display for one of the HS-PDSCHs illustrates that most of the cells capacity would be consumed by this high data rate HS-DSCH configuration. Realistic peak data rates are likely to be much lower than 14 Mbps. As an example, the most stringent single link conformance test requirement for the UE in Rel-5 is based on a five-code QPSK channel with nominal data rate of 1.6 Mbps. The required throughput is 1.269 Mbps. The most stringent closed loop diversity conformance test requirement for the UE in Rel-5 is based on a four-code 16QAM channel with nominal data rate of 2.332 Mbps. The required throughput in this case is 1.5 Mbps. In both these test cases the cell needs to be 10 dB above the noise and the UE consumes HALF the cell power.

25

Hybrid-ARQ
Hybrid
Feed forward error correction, such as convolutional coding and turbo coding ARQ : Feedback error correction.

ARQ : Automatic Repeat reQuest

Transmitter 1 2 3
AC K

4
NA CK

5
AC K

2
AC K

6
NA CK

7
AC K

Receiver

2
ERROR

5
ERROR

ACK/NACK control : Layer 1 Packet Order Scheduling : Layer 2 (MAC)

Concepts of HSDPA rev.8.1 November 9th, 2005 Agilent Copyright Page 26

Hybrid-Automatic Repeat reQuest (HARQ) is a combination of feed forward error correction (FEC) and ARQ methods that saves information from previous failed attempts to be used in future decoding. It is based on an asynchronous downlink and a synchronous uplink. The uplink reports either ACK or NACK upon reception and decoding of a block. HARQ is an implicit link adaptation technique. Whereas in AMC explicit C/I measurements or similar measurements are used to set the modulation and coding format, in HARQ, link layer acknowledgements are used for re-transmission decisions. AMC provides the coarse data rate selection, while HARQ provides for fine data rate adjustment based on channel conditions.

26

Hybrid-ARQ Retransmission
Upon Receiving a NACK Data Can be Retransmitted Number of Information Bits (NINFO) Retransmitted are the Same as the Initial Transmission Modulation Scheme, Channelization Code Set and the Transmission Power May be Different From Original Transmission Number of Channel Bits (NCH) May Be Different Compared to Original Transmission Even if NCH is the Same as the Original Transmission the Set of Bits Retransmitted May be Different as Compared to the Original Set

Ref : 25.858 7.1

Concepts of HSDPA rev.8.1 November 9th, 2005

Agilent Copyright

Page 27

For a retransmission, the transport-block set is the same as for the initial transmission. This means that, for a retransmission, the number of information bits NINFO to be transmitted is the same as for the initial transmission.

For a retransmission the modulation scheme and the channelization-code set, including the size of the channelization-code set, and the transmission power, may be different compared to the initial transmission. This means that, for a retransmission, the number of available channel bits Nch may differ compared to the initial transmission. The channel bits are the bits actually transmitted over the air. Even if the number of available channel bits Nch is the same, the set of channel bits may be different for the retransmission compared to the initial transmission.

27

Soft Combining Strategies

Soft Combining Uses Information From Both Original Transmission and Retransmissions Chase Combining (CC) Identical Version of Erroneously Detected Packet is Transmitted Received Copies are Combined by the Decoder Based on SNR Prior to Decoding Incremental Redundancy (IR) Additional Redundant Information is Incrementally Transmitted

Ref : 25.858 7.1

Concepts of HSDPA rev.8.1 November 9th, 2005 Agilent Copyright Page 28

Soft combining is used in HSDPA to minimize the need for further repeat requests. Two choices of HARQ combining schemes are possible: Chase Combining (CC) and Incremental Redundancy (IR). CC corresponds to defining or using only a single redundancy version. In IR a different set of bits is retransmitted and combined with the original set, hence increasing the amount of redundant data and the likelihood of recovering from the errors introduced on the air.

28

Incremental Redundancy
Transmitted data: UE Redundancy buffer:

Sorry. Didnt hear you? (NACK)

n e n l e n n In re en al re un an y

Almost got it! (NACK)

c m t _ d d c Incremental_redundancy

OK, heard you that time! (ACK)

Dowt trwnsmwt awy mowe dwtw twan ywu havw to

Dont transmit any more data than you have to
Concepts of HSDPA rev.8.1 November 9th, 2005 Agilent Copyright Page 29

The idea behind incremental redundancy is to transmit as few bits as possible by heavily puncturing the data. At each stage of retransmission, the effective throughput decreases. The slide depicts an analogy of how incremental redundancy works. In this example, if the first transmission had been heard there would be a 3x gain. By the time the second transmission happens, the potential advantage is reduced to 1.5 times. Since all three transmission are needed to decode the message, there is no gain from having used incremental redundancy. The benefit of incremental redundancy over chase combining depends on the effective code rate. A low effective code rate implies that there is not much puncturing applied, to begin with. So incremental redundancy does not provide a significant advantage over chase combining in that case, because there is little redundancy that can be added to the received data.

29

Inter-TTI Interval
The HSDPA Transmission Time Interval (TTI) is always 2 ms (=1 subframe) Inter-TTI interval refers to the number of TTIs (or subframes) between transmissions to the same UE All UE must be capable of receiving transmissions every 3rd sub-frame which is an Inter-TTI interval of 3 More capable UE can handle an Inter-TTI interval of 2 or 1
HS-PDSCH(s) Data DTX DTX Data DTX DTX Data DTX DTX

Inter-TTI = 3 HS-PDSCH(s) Data DTX Data DTX Data DTX Data DTX Data

Inter-TTI = 2
Concepts of HSDPA rev.8.1 November 9th, 2005 Agilent Copyright Page 30

The HSDPA Transmission Time Interval (TTI) is always 2 ms, so it is equivalent to 1 subframe. The Inter-TTI interval refers to the number of TTIs (or subframes) between transmissions to the same UE. The minimum inter-TTI interval a UE must support is either 1,2 or 3. This value depends on the HS-DSCH category (defined next). UE with a minimum inter-TTI interval of 1 should be capable of receiving data every sub-frame, a value of 2 means data every other sub-frame, and 3 is every third sub-frame. Therefore, the most capable UEs have a minimum inter-TTI interval of 1.

30

HARQ Processes (1)

If only 1 HARQ process, inter-TTI interval is at least 6 because arrival of ACK/NACK takes about 5 sub-frames Multiple Independent HARQ Processes Per User May Run in Parallel
Data Rate to User is Increased Available System Capacity is Not Wasted
Sub-frame # 1 P=0 NDI=0 2 DTX 3 DTX 4 P=1 NDI=0 DTX 5 DTX 6 DTX 7 P=0 NDI=1 DTX 8 DTX 9 DTX 10 P=1 NDI=0 DTX 11 DTX

Process #0 Process #1
12 DTX 13 P=0 NDI=0 DTX

HS-SCCH

HS-PDSCH(s)
2 slots

Data

DTX

Data

DTX

Data

DTX

Data

DTX

Inter-TTI = 3
ACK ACK DTX DTX DTX DTX DTX DTX
Page 31

CQI

CQI

CQI

NACK

HS-DPCCH

UE Processing delay ~2.5 sub-frames (7.5 slots)

CQI

CQI

CQI

CQI

CQI

Concepts of HSDPA rev.8.1 November 9th, 2005

Agilent Copyright

Data is not retransmitted until the receipt of the ACK or NACK for that data. A HARQ process can only be used for one block of data at a time. It takes around 5 subframes of time to get the associated ACK/NACK response at the BTS as to if the UE correctly received the BTS transmission, so the inter-TTI interval for a single HARQ process is at least 6. In order not to waste the latency time between transmission of the block and reception of the ACK/NACK response, multiple independent HARQ processes can be used to transmit more blocks of data. This increases the data rate to a specific UE. The slide shows 2 HARQ processes and an inter-TTI of 3. Each HARQ process is assigned an identifier (P=0 or 1 in the slide), which is used to couple the processes in the transmitter and the receiver. The new data indicator (NDI) is used to distinguish between data blocks. The HS-SCCH indicates new data being transmitted by toggling the NDI value between 0 and 1 within the same process (see 25.321 11.6.1.3). In this example, Process #1 receives a NACK, which results in a retransmission of a packet (the NDI value stays as a 0).

31

HARQ Processes (2)

Throughput can be increased by reducing the inter-TTI period, which allows the interleaving of more HARQ processes A UE capable of inter-TTI = 3 supports 2 HARQ processes (or less) A UE capable of inter-TTI = 2 supports 3 HARQ processes (or less) A UE capable of inter-TTI = 1 supports the maximum 6 processes (or less) The BTS chooses any number processes up to the max the UE class handles
Sub-frame # 1 Control 2 DTX 3 Control 4 DTX 5 Control 6 DTX 7 Control 8 DTX 9 Control 10 DTX 11 Control 12 DTX 13 Control

HS-SCCH

HS-PDSCH(s)
2 slots

Data

DTX

Data

DTX

Data

DTX

Data

DTX

Data

DTX

Data

DTX

Inter-TTI = 2
DTX DTX DTX ACK ACK ACK ACK ACK
Page 32

Concepts of HSDPA rev.8.1 November 9th, 2005

Agilent Copyright

~2.5 sub-frames (7.5 slots)

Throughput can be increased by reducing the inter-TTI interval, which allows the interleaving of more HARQ processes. With an inter-TTI interval=2 and 3 HARQ processes, the throughput is 1.5 times the throughput with an inter-TTI interval of 3 and 2 HARQ processes. A UE with minimum inter-TTI interval of 3 must be able to support 2 HARQ processes. A UE with minimum inter-TTI interval of 2 must be able to support 3 HARQ processes. A UE with minimum inter-TTI interval of 1 must be able to support 6 HARQ processes, which are the maximum number of processes that can be transmitted in parallel. Soft memory is partitioned across HARQ processes. The BTS chooses any number of processes up to the maximum that the UE class can handle.

DTX

HS-DPCCH

Processing delay

CQI

CQI

CQI

CQI

CQI

CQI

CQI

CQI

32

UE Capability Classes
HS-DSCH category (FDD) Maximum number of HSDSCH codes received 5 5 5 5 5 5 10 10 15 15 5 5 Minimum inter-TTI interval 3 3 2 2 1 1 1 1 1 1 2 1 Maximum number of bits of an HS-DSCH transport block received within an HSDSCH TTI 7298 7298 7298 7298 7298 7298 14411 14411 20251 27952 3630 3630 Total number of soft channel bits 19200 28800 28800 38400 57600 67200 115200 134400 172800 172800 14400 28800

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

UEs of Categories 11 and 12 support QPSK only.

Concepts of HSDPA rev.8.1 November 9th, 2005 Agilent Copyright

Ref : 25.306 4.5.3 and table 5.1a

Page 33

The table from the specifications defining the UE categories (HS-DSCH physical layer categories (FDD) from 25.306 (table 5.1a)) is shown above. The main parameters used to define the UE physical layer capabilities are the following: The maximum number of HS-DSCH codes supported The minimum inter-TTI interval. The maximum number of HS-DSCH transport channel bits (or transport block size) that can be received within a single TTI. The maximum number of soft channel bits over all the HARQ processes Support for 16QAM, or QPSK only The UE must be capable of receiving 4 parallel HS-SCCHs.

33

MAC Layer Scheduling Functionality in Node-B

Packet Scheduling Functionality located in MAC Layer in Node-B (BTS) instead of in Network Controller
Based on channel quality, UE capability, Quality of Service (QoS) class, power/code availability Fast scheduling possible because of functionality closer to air interface and short frame length Fast adaptation to channel variation (AMC) Fast HARQ retransmissions

Concepts of HSDPA rev.8.1 November 9th, 2005

Agilent Copyright

Page 34

In order to support fast packet transfer, the packet scheduling functionality is added to the MAC Layer in the BTS. It is possible to schedule according to radio channel conditions (based on the CQI from all the UEs involved) as well as the amount of data to be transmitted. Scheduling can become computationally complex, depending on the algorithm used. The maximum throughput comes from serving the UE(s) with the best radio channel conditions but practically some degree of fairness is needed. The actual throughput is heavily dependent on the packet scheduling algorithm used. Scheduling, modulation and coding adaptation, and HARQ retransmissions are fast because they are performed as close to the air interface as possible and because a short frame length is used. Fast scheduling makes it possible to track the fast channel variations.

34

HSDPA Test Models and Reference Channels

Test Model 5 for BTS transmitter testing:
Ref : 25.141 6.1.1.4A

Use to test EVM for BTS supporting HSDPA w/ 16QAM modulation Includes 8, 4, or 2 HS-PDSCHs depending on BTS capability 5 (6 for Release 6) Downlink Fixed Reference Channels (FRC) specified for UE receiver and HSDPA performance requirements testing FRC H-Set 1 for UE of HS-DSCH category 1 and 2 FRC H-Set 2 for UE of HS-DSCH category 3 and 4 FRC H-Set 3 for UE of HS-DSCH category 5 and 6 FRC H-Set 4 for UE of HS-DSCH category 11 FRC H-Set 5 for UE of HS-DSCH category 12 FRC H-Set 6 for UE of HS-DSCH category 7 and 8
Concepts of HSDPA rev.8.1 November 9th, 2005 Agilent Copyright Page 35

Ref : 25.101 9

A new test model, Test model 5, has been added to the conformance specifications as the test configuration for BTS transmission testing. Similar to Reference Measurement Channels (RMC) in 3GPP Release 99, HSDPA defines FRCs. The different FRCs are used for UE receiver and performance requirements tests, depending on the UE category.

35

FRC H-Set 3
Parameter Nominal Avg. Inf. Bit Rate Inter-TTI Distance Number of HARQ Processes Information Bit Payload ( N INF ) Number Code Blocks Binary Channel Bits Per TTI Total Available SMLs,in UE Number of SMLs per HARQ Proc. Coding Rate Number of Physical Channel Codes Modulation Unit kbps TTIs Processes Bits Blocks Bits SMLs SMLs Codes Value 1601 2332 1 1 6 6 3202 4664 1 1 4800 7680 57600 57600 9600 9600 0.67 0.61 5 4 QPSK 16QAM

Ref: 25 101 A.7.1.3

Concepts of HSDPA rev.8.1 November 9th, 2005

Agilent Copyright

Page 36

As an example, the table from the specifications for the FRC H-Set 3 is shown above.

36

Coding FRC H-Set 3

Inf. Bit Payload CRC Addition Code Block Segmentation Turbo-Encoding (R=1/3) 1st Rate Matching RV Selection Physical Channel Segmentation
4800 3202 3202 3226 9678 12 Tail Bits 24 CRC

(QPSK)

9600

960

Inf. Bit Payload CRC Addition Code Block Segmentation Turbo-Encoding (R=1/3) 1st Rate Matching RV Selection Physical Channel Segmentation

4664 4664 4688 14064 24 CRC

(16QAM)

12 Tail Bits

9600 7680

1920

Ref: 25.101 A.7.1.3

Page 37

Concepts of HSDPA rev.8.1 November 9th, 2005

Agilent Copyright

The figure above corresponds to the coding for FRC H-Set 3.

37

What is next? Enhanced Uplink Dedicated Channel (EUDCH)

Also known as High Speed Uplink Packet Access (HSUPA) Targeted for Release 6 Increases uplink data rates, improves coverage, and reduces delay Supports multimedia services (packet based): video conferencing Employs techniques similar to HSDPA:
AMC Addition of 16QAM HARQ Scheduling at Node-B Shorter frame size (TTI)
Concepts of HSDPA rev.8.1 November 9th, 2005 Agilent Copyright Page 38

The enhanced uplink dedicated channel (EUDCH), more commonly known as HSUPA, is targeted for Release 6 of the 3GPP specifications. It is supposed to provide uplink data rates similar to those provided by HSDPA in the downlink, improve coverage, and reduce end-user delay. Its main application are packet based services that require high throughput in both the downlink and the uplink, such as video conferencing. The techniques used for EUDCH are anticipated to be similar to those used for HSDPA: AMC, HARQ, addition of 16QAM, scheduling controlled at the Node-B, and shorter frame size, among others.

38

What is next? Multiple Input Multiple Output (MIMO) Antennas

Targeted for Release 7 Increases data rates and capacity Theoretical capacity increases linearly with the number of Tx/Rx antennas
Combined Tx 1, Tx 2 signals

Tx 1

Rx 1

Tx 1 Data

DSP
Tx 2
Concepts of HSDPA rev.8.1 November 9th, 2005 Agilent Copyright

Rx 2

Tx 2 Data
Page 39

Multiple Input Multiple Output (MIMO) antenna systems get their name from the use of multiple antennas at the transmitter (input to the channel) and multiple antennas at the receiver (output of the channel). Unlike other smart-antenna systems, in this system the data stream is split among the multiple antennas, so each antenna is transmitting a unique data stream. The primary goal is to increase the data rate and the spectral efficiency. MIMO systems rely on multipath to ensure that each receiver sees a different linear combination of transmit signals. When that condition is met, the receiver can estimate the channel characteristics from each and every transmit antenna, to each and every receive antenna and then use that information to isolate each transmit signal. For example, in a system with two transmit antennas and two receive antennas, four channel responses would need to be estimated. In theory, this system can increase peak data rates by the number of transmit/receive pairs. So a two antenna system could reach 2 times the peak data rate of Release 5 HSDPA. The incorporation of MIMO antennas are expected to be part of Release 7 of the 3GPP specifications.

39

Which of the following statements about AMC and HARQ in HSDPA are correct?
The modulation and coding can be adjusted every subframe Both CQI and HARQ ACK/NACK fields provide information on channel quality The only parameters that determine the data rate within a subframe are the modulation scheme and the number of HSPDSCHs used. Users with better channel conditions are typically assigned higher effective code rates and higher order modulation schemes.
Agilent Copyright Page 40

Concepts of HSDPA rev.8.1 November 9th, 2005

Information about the Q&A (review slides 23, 24, and 26 for further information):
Which of the following statements about AMC and HARQ in HSDPA are correct? The modulation and coding can be adjusted every subframe True. AMC allows the modulation, coding, and number of HS-PDSCHs to be adjusted every subframe (2 ms). (review slide 23 and 24) Both CQI and HARQ ACK/NACK fields provide information on channel quality True. The CQI is an explicit indicator of channel quality, while the HARQ ACK/NACK information is an implicit indicator of channel quality (if the UE sends a NACK, it means that the block was not well received/decoded, which, ultimately means that the channel quality was worse than expected, assuming that the UE is performing as expected). (review slide 26) The only parameters that determine the data rate within a subframe are the modulation scheme and the number of HS-PDSCHs used. False, the nominal data rate (data rate within a subframe) is defined by the transport block size. The transport block size, modulation scheme and number of HS-PDSCHs determine the effective code rate. In other words, different transport block sizes (or nominal data rates) are possible because different effective code rates, modulation schemes, and number of HS-PDSCHs can be used. (review slide 24) Users with better channel conditions are typically assigned higher effective code rates and higher order modulation schemes. True. This is the basic principle behind AMC. Higher effective code rates (within a subframe) means less coding redundancy (within a subframe), so less margin for error in the reception. Higher order modulation schemes also have less margin for error. Therefore, it only makes sense to use high effective code rates and high order modulation schemes when we have very good channel conditions, which minimize the probability of error. (review slide 23)

40

For an HSDPA HARQ retransmission...

The number of information bits may be different compared to the initial transmission The set of channel bits must be different as compared to the original set The transmission power may be different from the original transmission The modulation scheme and channelization code set might be different from the original transmission

Concepts of HSDPA rev.8.1 November 9th, 2005

Agilent Copyright

Page 41

Information about the Q&A (review slide 27 for further information): For an HSDPA HARQ retransmission...
The number of information bits may be different compared to the initial transmission False. The information (original data bits in the transport block) that we want to transmit is always the same in the original transmissions and subsequent retransmissions. The set of channel bits must be different as compared to the original set False. The set of physical channel bits can be different, but it can also be the same, depending on the combining scheme used. The transmission power may be different from the original transmission True. The transmission power can be changed from one transmission to another one. The modulation scheme and channelization code set might be different from the original transmission. True. The modulation scheme and channelization code set (including the number of HSPDSCHs) can be changed for a retransmission. So, AMC can be applied even if we are retransmitting a block, except that the transport block size has to be the same as in the original transmission.

41

 For a retransmission, the redundancy version and constellation version are different from the original transmission. For a retransmission, the redundancy version and constellation version are the same from the original transmission. The idea is to transmit as few bits as possible by heavily puncturing the data. Data throughput decreases as the number of retransmissions increases.
Page 42

Which of the following statements about incremental redundancy are correct?

Concepts of HSDPA rev.8.1 November 9th, 2005

Agilent Copyright

Information about the Q&A (review slide 29 for further information):

Which of the following statements about incremental redundancy are correct? For a retransmission, the redundancy version and constellation version are different from the original transmission. True. In Incremental Redundancy, different puncturing schemes are used in the retransmissions to increase the redundancy of the data received, and minimize the probability of error. Different constellation re-arrangements for 16 QAM are also possible. For a retransmission, the redundancy version and constellation version are the same from the original transmission. False. This is true for Chase Combining but not for Incremental Redundancy. The idea is to transmit as few bits as possible by heavily puncturing the data. True. In other words, the idea is not to transmit any more data than you have to. Data throughput decreases as the number of retransmissions increases. True. More retransmissions mean more time spent (or a larger total number of bits transmitted) trying to transmit/receive the same original information successfully, which is equivalent to lower throughput for the end-user.

42

What is the maximum number of HARQ processes that a UE with a minimum inter-TTI interval of 3 must support?
6 3 2 1

Concepts of HSDPA rev.8.1 November 9th, 2005

Agilent Copyright

Page 43

Information about the Q&A (review slides 31, 32, and 33 for further information):
What is the maximum number of HARQ processes that a UE with a minimum inter-TTI interval of 3 must support? 6 False. The only UEs that can handle 6 parallel HARQ processes are UEs with minimum inter-TTI interval of 1. 3 False. UEs with minimum inter-TTI interval of 2 must be able to handle 3 HARQ processes (they can also handle fewer than 3). UEs with minimum inter-TTI interval of 1 can also handle 3 processes (they can handle any number of parallel processes, from 1 up to 6 processes). 2 True. UEs with minimum inter-TTI interval of 3 must be able to handle 2 HARQ processes (they can also handle 1 process). These are the lower performing UEs. All the UEs should be able to handle 2 parallel processes (or more, for UEs with minimum inter-TTI interval of 1 or 2). 1 False. All the UEs should be able to handle 1 process, but they should be able to handle more than that, depending on their capability.

43

In HSDPA, data throughput for a user may depend on which of the following?
The channel conditions for that user The channel conditions for all users The amount of data to be transmitted to the different users The packet scheduling algorithm The UE capability class

Concepts of HSDPA rev.8.1 November 9th, 2005

Agilent Copyright

Page 44

Information about the Q&A (review slide 34 for further information):

In HSDPA, data throughput for a user may depend on which of the following? The channel conditions for that user True. AMC allows the modulation and coding to vary according to the instantaneous channel conditions. So, the data rate for a user can change from subframe to subframe. The data rate used affects the overall throughput for that user. The channel conditions will also affect the number of retransmissions required to successfully decode a block, which also affects the overall throughput for that user. The channel conditions for all users True. The packet scheduling algorithm must take into account the channel conditions to all users to figure out how to distribute resources. So, the channel conditions for the different users affect the overall throughput for the single user. The amount of data to be transmitted to the different users True. The packet scheduling algorithm must take into account the amount of data to be transmitted to the different users to figure out how to distribute resources. So, the amount of data for the different users affects the overall throughput for the single user. The packet scheduling algorithm True. The packet scheduling methodology and the actual algorithm have a direct effect on the overall throughput for the single user. The UE capability class True. Higher performing UEs (minimum inter-TTI interval of 1, capability to receive more HSPDSCHs, higher block transport size, etc) can handle higher data rates.

44

Agilent Solutions for HSDPA

Design Simulation & Verification

Agilent ADS Communications Systems Designer and W-CDMA Design Library

Agilent 89601A Vector Signal Analysis Software

DUT Agilent E4438C ESG Vector Signal Generator Signal Studio for HSDPA over W-CDMA
Concepts of HSDPA rev.8.1 November 9th, 2005 Agilent Copyright

Agilent E4440A PSA Spectrum Analyzer (or 89600 Vector Signal Analysis hardware or E4406A VSA Transmitter Tester or ESA-E Series Spectrum Analyzers or Infiniium Oscilloscopes)
Page 45

This slide shows the recommended process for simulating and verifying the HSDPA performance of a W-CDMA BTS or UE device or component design. First, the whole system is created and simulated using the Agilent ADS Communications Systems Designer and W-CDMA Design Library. The simulated signal is analyzed using the Agilent 89601A Vector Signal Analysis Software, which is dynamically linked with the ADS simulation tool on the PC platform, and the design parameters are adjusted accordingly. Once the simulation measurements provide results within the expected value range, the different hardware prototype sections can be built. The signal that is used as a stimulus for each section in the simulation is transferred to an Agilent E4438C ESG Vector Signal Generator. This generator can then provide a stimulus signal for the hardware prototype on the bench with the same characteristics as the one used in the simulation., Signal Studio for HSDPA over W-CDMA can also be used as an alternative to ADS to configure fully coded downlink and uplink HSDPA channels, when intermediate access to the coding layers is not required. (Note: The fully coded DPCH channels that are included in Signal Studio lack the advanced feature set included in option 400 (W-CDMA personality), i.e. compressed mode, etc). Each prototype section is verified using the Agilent E4440 PSA Spectrum Analyzer. The D.U.T. output test signal can also be read from the PSA back into the ADS simulation environment using the Agilent 89601A Vector Signal Analyzer (VSA) software, which also offers signal source capability, to read signals from the testbench/PSA back into ADS. This means that the effects of the DUT hardware prototype in the overall system can be assessed even before all the sections are built, which simplifies the design rework cycles. Once all the hardware prototype sections have been through this cycle, they are physically integrated and the whole transmitter system is verified on the bench using the Agilent E4440A. The 89601A VSA software runs inside of ADS with the same powerful post-processing algorithms used by the PSA on the bench, so results between simulated and test correlate. The 89600 VSA hardware, the E4406A Vector Signal Analyzer (which is mostly used in manufacturing), the ESA-E Series Spectrum Analyzers (general purpose SA), or the Infiniium Oscilloscopes, can also be used in place of the E4440A PSA.

45

Agilent Solutions for HSDPA

Baseband Design Simulation & Verification

Agilent ADS Communications Systems Designer and W-CDMA Design Library

Agilent 89601A Vector Signal Analysis Software Digital BB DUT Agilent 1680, 1690, or 16900 logic analyzers and pattern generators
Page 46

Signal Studio for HSDPA over W-CDMA

Concepts of HSDPA rev.8.1 November 9th, 2005

Agilent E4438C ESG Baseband Studio Vector Signal Digital Signal Generator Interface Module
Agilent Copyright

For DUTs with digital inputs and/or outputs, a similar process applies. The Baseband Studio Digital Signal Interface Module (DSIM) provides digital inputs and outputs for the E4438C ESG Vector Signal Generator. The DSIM is also available for direct connectivity with a PC platform, so ADS or custom signals can be transferred directly to a DUT with digital inputs without the need of a signal generator. Direct connectivity with the 89601A vector signal analysis software is also provided, for DUTs with digital outputs. As an alternative for DUTs with digital outputs, Agilent 1680, 1690, and 16900 logic analyzers can be used to capture the digital signal and transfer it to ADS or the 89601A software for further analysis. The 16900 family also has a pattern generator option with connectivity to ADS that can also be used to provide digital stimulus to DUTs with digital inputs.

46

ADS W-CDMA Design Library

Test Model 5 in ADS 2003C

Select 6,14, or 30 DPCHs

3GPP
...
TestModel5

Q
Mod_RC

VSA

I RF_ModulatorRC IF_Mod SymbolRate=3.84e6 MirrorSpectrum=NO VSA_89600_Sink V1

3GPPFDD_TestModel5 CxToRect G1 C1 SpecVerHSDPA=Version 09-03 ScrambleCode=0 ScrambleOffset=0 ScrambleType=Normal SCCPCH_SltFmt=slot format 0 OutputMode=Ramp Seed=0 DF DPCHSet=6 DPCH DF

Concepts of HSDPA rev.8.1 November 9th, 2005

Agilent Copyright

Page 47

The W-CDMA Design Library in ADS 2003C includes Test Model 5. As mentioned earlier, test model 5 is the channel configuration defined in the specifications to verify the modulation quality performance of the HSDPA BTS transmitter or component.

47

ADS W-CDMA Design Library

HSDPA Coded Uplink Source and Receiver in ADS 2004A

Concepts of HSDPA rev.8.1 November 9th, 2005

Agilent Copyright

Page 48

The W-CDMA Design Library in ADS 2004A includes both an HSDPA coded uplink source and an HSDPA coded uplink receiver. The BTS receiver is particularly useful during UE performance requirement simulation and verification to decode the ACK/NACK and CQI response from the UE, which is necessary to calculate BLER and other performance requirement metrics.

48

Agilent Solutions for HSDPA

-Design VerificationPre-defined test model 5 16QAM demod bits support Modulation accuracy with HSDPA signal AMC support in code domain

Agilent E4440A PSA Spectrum Analyzer

Agilent E4438C ESG Vector Signal Generator with Signal Studio for HSDPA over W-CDMA (option 418)

Easy to use graphical user interface Agilent E4406A VSA Transmitter Tester Real-time encoded CRC for HS-DSCH DL pattern support for UE ACK/NACK and CQI responses Flexible uplink ACK/NACK & CQI patterns Pre-defined RMC and FRC channels Open loop transmit diversity support Agilent 89600 Series Vector Signal Analyzers Pre-defined test model 5 (option 400)
Concepts of HSDPA rev.8.1 November 9th, 2005 Agilent Copyright Page 49

The Agilent solutions for signal generation and analysis of HSDPA BTS and UE are shown above. The Agilent E4438C ESG Vector Signal Generator with Signal Studio for HSDPA over W-CDMA (option 418) enables you to perform receiver testing on W-CDMA BTS and UE that use the HSDPA channels. Note that Signal Studio is not suitable for amplifier testing of signals with the HSDPA channels. Instead, the E4438C ESG Vector Signal Generator with the 3GPP W-CDMA FDD personality (option 400) should be used for this purpose. This option is intended for testing both amplifiers and receivers for the primary W-CDMA FDD communications channels. It also contains Test Model 5 for testing amplifiers with the HSDPA channels. It is not suitable for BER/BLER testing of HSDPA receivers. The HSDPA measurement personality (option 210) of the Agilent E4440A PSA Spectrum Analyzer and the E4406A VSA Transmitter Tester enables HSDPA signal analysis for modulation quality very easily. And the Agilent 89600 Series Vector Signal Analyzers with option B7N, 3G modulation analysis, also provides transmitter measurements for signals with HSDPA channels, such as code domain power and composite EVM.

49

ESG with Signal Studio for HSDPA over W-CDMA UE test features enable BLER testing
Common Channels Multicode HSDPA Channels 4 HS-PDSCH 4 HS-SCCH 16 OCNS Channels Set AWGN

Transport layer config 16QAM or QPSK Modulation

Agilent Copyright Page 50

Concepts of HSDPA rev.8.1 November 9th, 2005

The figure above shows some of the features of the Signal Studio for HSDPA over W-CDMA software for UE testing. The most important feature for testing mobiles is the ability to create downlink transport and physical layer coded HSDPA signals for performing BLER analysis. The transport layer coding can be turned off for troubleshooting applications. The software enables the user to generate up to 15 multicodes, and select either QPSK or 16QAM for the downlink HSDPA transmissions. This is useful for testing the demodulation and decoding algorithms in the baseband section. The next feature is the ability to configure the W-CDMA common control channels. Although these are not part of the HSDPA capability, they are needed to establish a radio link between the mobile and the ESG. Another feature is the ability to generate up to 16 OCNS channels. These channels simulate the interference observed by the mobile due to the BTS simultaneously transmitting to other users in a particular cell. A related feature is the AWGN capability. The AWGN capability simulates the interference caused by other transmitters (TV stations, radio stations, other cellular communications, etc) in the vicinity of the cell area. One final feature is generating up to 4 downlink HSDPA data channels and the corresponding HSDPA control channels. This is useful for testing whether the mobile can correctly identify the appropriate HSDPA channel to demodulate and decode. Signal Studio for HSDPA over W-CDMA also provides the ability to configure two ESGs for open loop diversity (in a different part of the user interface).

50

ESG with Signal Studio for HSDPA over W-CDMA HARQ and AMC support for UE performance testing
Test Adaptive Modulation Coding (AMC) capability
Configure CQI pattern and UE category Time varying streams of 16QAM and QPSK modulation

Test Hybrid Automatic Repeat Request (HARQ) capability

Configure ACK/NACK pattern Configure RV parameter
ESG emulates BTS Tx System Each packet is created with different coding parameters

CRC result

Create ACK/NACK or CQI patterns

Concepts of HSDPA rev.8.1 November 9th, 2005 Agilent Copyright Page 51

Signal Studio for HSDPA over W-CDMA enables functional and performance testing of the AMC and HARQ features associated with HSDPA transmissions. Each packet is coded with different parameters depending upon the CQI or ACK/NACK information. The feature works in an open loop mode where the CQI or ACK/NACK information is simulated through the use of a pattern configured by the user. These CQI or ACK/NACK patterns can be up to 1280 subframes in length giving the user the capability to create sophisticated AMC or HARQ scenarios. Note that conformance testing is not possible with this solution, since a closed feedback control loop for the HARQ and AMC parameters is required.

51

ESG with Signal Studio for HSDPA over W-CDMA BTS test features enable functional testing

CQI pattern

ACK/NACK pattern

Concepts of HSDPA rev.8.1 November 9th, 2005

Agilent Copyright

Page 52

The figure above shows some of the features of the Signal Studio for HSDPA over W-CDMA software for BTS testing. The most important feature with regard to HSDPA BTS testing is the ability to create ACK/NACK and CQI patterns to determine the BTS ability to correctly respond to different scenarios. Another feature is the ability to create the basic W-CDMA signal for use in establishing a radio link between the ESG and the BTS.

52

HSDPA Personality in Signal Analyzers

Pre-defined Test Model 5
Symbol Boundary: Auto TM5/8xHS-PDSCH

AUTO active detection fails because code channel here is set to 30dB
Concepts of HSDPA rev.8.1 November 9th, 2005 Agilent Copyright

With Predefined Test Model, code channel is set to Active by force

Page 53

The following slides show some of the features of the HSDPA personality (option 210) in the PSA and E4406A VSA signal analyzers. By selecting a predefined test model the instrument is forced to despread and demodulate using the pre-defined parameters listed below: Modulation Scheme (QPSK or 16QAM) SF (Spreading Factor) Channelization Codes

53

HSDPA Personality in Signal Analyzers

Adaptive Modulation and Coding
Slot #0 Detected as QPSK correctly Move to slot #3 HS-PDSCH modulated in 16QAM

Concepts of HSDPA rev.8.1 November 9th, 2005

Agilent Copyright

Page 54

The HSDPA personality in the Agilent signal analyzers supports both QPSK and 16QAM modulation schemes and can detect them automatically. The modulation scheme detection has three modes: Auto: Automatically detects the modulation scheme (16QAM or QPSK) by slot-by-slot estimations. QPSK: Forces to detect as QPSK 16QAM: Forces to detect as 16QAM

Symbol Boundary : Auto (Detects active channels within Meas Offset/Meas Interval range). Mod Scheme : Auto (Detects modulation scheme only within Meas Offset/Meas Interval specified range).

54

HSDPA Personality in Signal Analyzers

16QAM Demodulated Bits Support
I1 1011 Q1 1010 1110 1111 1001 1000 1100 1101 0001 0000 0100 0101 0011 0010 0110 0111 I2 Q2 Q2

i b(I)

0 I1

1 Q1

2 I2

3 Q2

*From TS25.213 5.1

Binary Format

I2

Hexadecimal Format

Concepts of HSDPA rev.8.1 November 9th, 2005

Agilent Copyright

Page 55

When the code channel is detected as 16QAM modulated, the HS-PDSCH demodulated bits can be seen in binary or hexadecimal formats.

55

HSDPA Personality in Signal Analyzers

Excel VBA HS-SCCH Decoding Tool

HS-SCCH code number PSA/E4406A IP address UE ID New data indicator (NDI) HS-DSCH HARQ process Redundancy and HS-PDSCH(s) constellation Modulation Transport Block identifier version (RV) size scheme

HS-PDSCH(s) code set : code number offset and number of codes

Concepts of HSDPA rev.8.1 November 9th, 2005

Agilent Copyright

Page 56

An external excel visual basic for applications (VBA) macro can be used to extract the despread HS-SCCH signal from the PSA/E4406A and decode the HSDSCH control information carried by this channel: the HS-PDSCH(s) modulation scheme and code set, and the transport channel parameters: transport block size, HARQ process identifier, RV, and NDI. Note: The excel macro shown in this slide is an internal Agilent tool. Please contact your local AE if you are interested in having a copy.

56

HSDPA Personality in Signal Analyzers

Excel VBA HS-DPCCH Decoding Tool

HS-DPCCH code number and I/Q allocation

Subframe Delay HARQ ACK/NACK

Concepts of HSDPA rev.8.1 November 9th, 2005

CQI
Agilent Copyright Page 57

An external excel VBA macro can also be used to extract the despread HSDPCCH signal from the PSA/E4406A and decode the ACK/NACK and CQI fields. Note: The excel macro shown in this slide is an internal Agilent tool. Please contact your local AE if you are interested in having a copy.

57

HSDPA Personality in Signal Analyzers

Modulation Accuracy for HSDPA Signal
Correct EVM reading Signal with HSDPA channels

Concepts of HSDPA rev.8.1 November 9th, 2005

Agilent Copyright

Page 58

When the HSDPA capability is enabled, 16QAM modulated code channels can be correctly detected for the EVM measurement. Without the HSDPA option, the HS-PDSCHs with 16QAM modulation would be demodulated assuming a QPSK constellation, which would result in large errors.

58

8960 with HSDPA Test Application for UE

E1963A Key Features
Most complete 34.121 section 5 & 6 test coverage in one box (including standard W-CDMA tests) Flexibility in UE Control FDD Test Mode RB Test Mode Fully functional MAC-hs 12.2 kbps RMC + FRC H-Set 1 to 5

HSDPA Measurements

Receiver and performance tests:

BLER and t-put R ACK/NACK count and statDTX count Median CQI
Concepts of HSDPA rev.8.1 November 9th, 2005 Agilent Copyright

Transmitter tests:
Channel Power, ACLR, SEM, dynamic power analysis Flexible HSDPA triggering
Page 59

The E1963A is primarily intended for W-CDMA, HSDPA, UMTS production test engineers who are developing test plans to manufacture HSDPA, W-CDMA, UMTS devices, Agilents one-box test set provides the most complete test functionality for TS 34.121 transmitter (section 5) & receiver characteristics (section 6) tests. Fast measurements and options in UE connectivity enable flexibility in test plan development and the assurance that designs meet technology standards. The E1963A can also be used for design engineers allowing them to speed the validation of their designs allowing them to get their products to market in the least amount of time. Both FDD (Frequency Division Duplex) and RB (Radio Bearer) test modes are available. FDD provides a basic HSDPA forward physical channel signal, no protocol is sent to communicate to the UE. This is ideal for efficient parametric test needs focusing on the RF transmitter and Receiver. RB test mode provides the same parametric test capability but adds the protocol required to bring the call up. Other key features include a new and completely functional MAC-hs layer that provides ACK/NACK support along with analysis of the CQI values reported by the UE, and generation of the standard W-CDMA control channels and a dedicated W-CDMA channel configured as the 12.2 kbps RMC, in addition to the HSDPA channels. FRC H-Sets 1 through 5 are supported, both in their QPSK and 16QAM configurations. The transmitter measurements supported include Channel power, ACLR, and SEM, along with a graphical burst by burst display of the UEs output power. These measurements, when coupled with a new flexible trigger for HSDPA, provide industry leading transmitter measurement capabilities, allowing customers to isolate transmission trouble within three distinct zones within the subframe. The receiver and performance requirement metrics supported include a BLER and t-put R in kbps, ACK/NACK count, statDTX count, and median CQI, as shown in the measurement display in the slide.

59

GS-8800 Plus: UMTS RF Conformance System for UE

Pre-Conformance and Conformance
Signal Analysis (E4445A PSA) Out-of-Band Interference (E8257D PSG) In-Band Interference (E4438C #1 ESG) RF Comm. Test Set (8960 #1 OBT) RF Comm. Test Set (8960 #2 OBT) RF Comm. Test Set (8960 #3 OBT)
Concepts of HSDPA rev.8.1 November 9th, 2005

Advantage of GS-8800 Plus

GSM Filter Module (N1962A) Add-on RF Interface BB Processor (Master) In-Band Interference (E4438C #2 ESG) BB Processor (Slave)
Agilent Copyright Page 60

PC Controller ATT/SW Driver (87130A)

RF Interface (N1961A)

RF

DUT

RF Fading Simulator with AWGN

Power Supply (66311B)

GSM/GPRS/EGPRS support GCF Certification Upgrade of Standard system to Plus system GS-8800 is scalable from Super-Lite, Lite, Standard, Plus1 (Pre-conformance) and Plus2 (Full conformance) Future HSUPA support

GS-8800 Plus Additional Instruments

The GS-8800 is a multiple radio technology RF design verification system. W-CDMA, HSDPA, GSM, (E)GPRS, cdma2000, 1xEVDO and AMPS are supported on the same hardware and software platform. The GS-8800 plus is a newly introduced W-CDMA/HSDPA RF conformance system based on the GS-8800 standard system. The GS-8800 plus expands the test coverage of the GS-8800 standard system to support more 3GPP TS 34.121 UE RF conformance test cases in Release 5 by adding baseband processors and other instruments used in conformance testing (specified in the block diagram in this slide using dotted red line). There are two configurations available for the GS-8800 plus system. The first one is the plus1 system which is a W-CDMA/HSDPA RF pre-conformance system. Pre-conformance means that the test coverage is limited to test cases which use only one 8960 RF test set. The second configuration is the plus2 system which is a W-CDMA/HSDPA RF full conformance system. This system includes HSDPA open loop and closed loop diversity test cases in 3GPP TS 34.121, as shown in the next slide.

60

GS-8800 Plus: UMTS RF Conformance System for UE (cont.)

Section Test Description 5.2A 5.9A

GS-8800 TS 34.121 HSDPA Tests

Maximum Output Power with HS-DPCCH Spectrum emission mask with HS-DPCCH

Standard system Plus1 Plus system Plus2 Plus1

5.10A Adjacent Channel Leakage Power Ratio (ACLR) with HS-DPCCH 5.13.1A Error Vector Magnitude (EVM) with HS-DPCCH 6.3A 9.2.1 9.2.2 9.2.3 9.3.1 9.3.2 9.4 Maximum Input Level for HS-PDSCH Reception (16QAM) Demodulation of HS-DSCH - Single Link Performance Demodulation of HS-DSCH - Open Loop Diversity Performance Demodulation of HS-DSCH - Closed Loop Diversity Performance Reporting of Channel Quality Indicator - AWGN Propagation Conditions Reporting of Channel Quality Indicator - Fading Propagation Conditions HS-SCCH Detection Performance

Concepts of HSDPA rev.8.1 November 9th, 2005

Agilent Copyright

Page 61

The slides shows the HSDPA test cases supported by the GS-8800 standard, plus1, and plus2 systems.

61

Anite Solutions for HSDPA UE Signaling Test

Anite W-CDMA/HSDPA SAT
2-12 cell signaling test solution for protocol development using the 8960 with the Anite Baseband Processor (ABP) 3GPP Release 5 Programmers Toolset (PT) 3GPP Release 5 Conformance Toolset (CT) with Conformance Test Cases HSDPA MAC-hs/RLC and baseband control via graphical user interface (GUI)

Anite W-CDMA/HSDPA SAS

2-12 cell signaling test solution for Interoperability Test (IOT) using 8960 with ABP 3GPP Release 5 network simulation via GUI
ABP ABP

Concepts of HSDPA rev.8.1 November 9th, 2005

Agilent Copyright

Page 62

The Anite SAT Test System provides functional verification of HSDPA UEs, as well as a complete Release 5 development test system. The system comprises multiple test units (up to 12) with an industry-standard PC controller. The test units are connected to the PC controller via a LAN interface. The test unit is comprised of an Agilent 8960 wireless communications test set and an Anite add-on module (the Anite Baseband Processor ABP). Multiple test units provide support for simulation of multiple cells and to allow handover procedures to be tested. The RF paths are combined via a straightforward power splitter/combiner. Introduction of other channel impairments such as co-channel or adjacent channel interferers can be done externally using a directional coupler or power splitter/combiner. The Anite HSDPA test solution includes both MAC and RLC functionality in accordance with 3GPP specifications TS 25.321 and TS 25.322 respectively. The following components of the MAC architecture are implemented: MAC-hs manages flow control, scheduling and priority control, per-user HARQ processes and transport format/resource combination selection, and MAC-d implements transport channel switching, C/T multiplexing and flow control with transport channel routing to MAC-hs. RLC entities can be activated in transparent mode, unacknowledged mode or full acknowledged mode and routed to logical channels as needed. The Anite SATsystem can be controlled in three ways: A C/C++ programmable API. This allows the test system user to write test programs or simulation scenarios in C/C++ using Anites full Programmers Toolset (PT) for W-CDMA & HSDPA. In addition to the functions of the test system itself, the programmer has access to the underlying facilities of the host computer such as file and I/O systems. This, for example, allows programs to control external equipment, to control the UE under test or to store and retrieve large amounts of test data. A test case manager using Anites Conformance Toolset (CT) to run conformance test scripts for both WCDMA and HSDPA. For HSDPA, a graphical user interface (GUI) is also available to allow control via MAC-hs/RLC and/or baseband for development testing. The Anite SAS system is controlled by a GUI that allows channels to be activated and deactivated. Test configurations can also be stored and reloaded. It allows for creation of complex cell or network simulations to enable interoperability testing.

62

Summary
HSDPA uses new techniques such as AMC and HARQ to achieve high peak rates HSDPA improves spectral efficiency for downlink packet data traffic New channels are defined for the downlink and the uplink TDM is used in combination with CDM Modulation and coding change in accordance with instantaneous variations in channel conditions HARQ and soft combining provide further data rate adjustment based on channel conditions MAC layer scheduling functionality in Node-B provides fast scheduling Agilent provides solutions for HSDPA testing of BTS and UE
Concepts of HSDPA rev.8.1 November 9th, 2005 Agilent Copyright Page 63

As a summary, HSDPA is a new technology incorporated in W-CDMA Release 5 to increase data throughput and improve the efficiency of the system for downlink data traffic. Some of the main changes in HSDPA are the addition of new channels, the combination of TDM with CDM techniques, the use of adaptive modulation and coding and Hybrid ARQ techniques, and the move of the MAC layer scheduling to the Node-B. Agilent currently provides a wide range of HSDPA test solutions for BTS and UE that can be used from early design verification through manufacturing.

63

Reference Documents
(www.3gpp.org)
3GPP TS 25.101 V.6.9.0 3GPP TS 34.121 V.6.2.0
UE Radio Transmission and Reception (FDD) Terminal Conformance Specification. Radio Transmission and Reception (FDD) BS Radio Transmission and Reception (FDD) Physical Channels and Mapping of Transport Channels onto Physical Channels (FDD) Multiplexing and Channel Coding (FDD) Spreading and Modulation (FDD) Physical Layer Procedures (FDD)
Agilent Copyright Page 64

3GPP TS 25.306 V.5.c.0 3GPP TS 25.308 V.5.7.0

UE Radio Access Capabilities HSDPA Overall description Medium Access Control (MAC) Protocol Specification High Speed Downlink Packet Access Physical Layer Aspects High Speed Downlink Packet Access: lub/lur Protocol Aspects UTRA High Speed Downlink Packet Access

3GPP TS 25.141 V.6.b.0 3GPP TS 25.211 V.5.7.0

3GPP TS 25.321 V.5.c.0

3GPP TR 25.858 V.5.0.0 3GPP TR 25.877 V.5.1.0 3GPP TR 25.950 V.4.0.0

3GPP TS 25.212 V.5.a.0 3GPP TS 25.213 V.5.6.0

3GPP TS 25.214 V.5.b.0

Concepts of HSDPA rev.8.1 November 9th, 2005

A listing of some of the Release 5 specifications that relate to HSDPA and that were used when developing this presentation is shown above.

64

Related Information
Information on Agilents HSDPA solutions: www.agilent.com/find/HSDPA Agilent Application Note Concepts of High Speed Downlink Packet Access: Bringing Increased Throughput and Efficiency to W-CDMA (literature p/n 5989-2365EN) Agilent Application Note HSDPA RF Measurements for User Equipment (literature p/n 5989-4099EN) Agilent Application Note HSDPA in the Agilent Technologies 8960 Wireless Communications Test Set (literature p/n 5989-3444EN)

Concepts of HSDPA rev.8.1 November 9th, 2005

Agilent Copyright

Page 65

65