60118029 hsdpa-parameter-description-130219192019-phpapp01

RAN
HSDPA
Parameter Description
Issue 02
Date 2009-06-30

Copyright © Huawei Technologies Co., Ltd. 2009. All rights reserved.
No part of this document may be reproduced or transmitted in any form or by any means without prior
written consent of Huawei Technologies Co., Ltd.
Trademarks and Permissions
and other Huawei trademarks are trademarks of Huawei Technologies Co., Ltd.
All other trademarks and trade names mentioned in this document are the property of their respective
holders.
Notice
The purchased products, services and features are stipulated by the contract made between Huawei and
the customer. All or part of the products, services and features described in this document may not be
within the purchase scope or the usage scope. Unless otherwise specified in the contract, all statements,
information, and recommendations in this document are provided “AS IS” without warranties, guarantees
or representations of any kind, either express or implied.
The information in this document is subject to change without notice. Every effort has been made in the
preparation of this document to ensure accuracy of the contents, but all statements, information, and
recommendations in this document do not constitute the warranty of any kind, express or implied.
Huawei Technologies Co., Ltd.
Address: Huawei Industrial Base
Bantian, Longgang
Shenzhen 518129
People's Republic of China
Website: http://www.huawei.com
Email: support@huawei.com
Huawei Proprietary and Confidential
Copyright © Huawei Technologies Co., Ltd

RAN
HSDPA Contents
Issue 02 (2009-06-30) Huawei Proprietary and Confidential
iii
Contents
1 Introduction to This Document...............................................................................................1-1
1.1 Scope.............................................................................................................................................................1-1
1.2 Intended Audience.........................................................................................................................................1-1
1.3 Change History..............................................................................................................................................1-1
2 Overview of HSDPA .................................................................................................................2-1
2.1 General Principles of HSDPA .......................................................................................................................2-1
2.2 HSDPA Channels ..........................................................................................................................................2-2
2.2.1 HS-DSCH and HS-PDSCH .................................................................................................................2-3
2.2.2 HS-SCCH.............................................................................................................................................2-3
2.2.3 HS-DPCCH..........................................................................................................................................2-3
2.2.4 DPCCH and DPCH/F-DPCH...............................................................................................................2-4
2.3 Impact of HSDPA on NEs.............................................................................................................................2-4
2.4 HSDPA Functions .........................................................................................................................................2-4
2.4.1 HSDPA Control Plane Functions .........................................................................................................2-4
2.4.2 HSDPA User Plane Functions..............................................................................................................2-6
3 Control Plane ..............................................................................................................................3-1
3.1 Bearer Mapping.............................................................................................................................................3-1
3.2 Access Control ..............................................................................................................................................3-2
3.3 Mobility Management...................................................................................................................................3-2
3.4 Channel Switching ........................................................................................................................................3-3
3.5 Load Control .................................................................................................................................................3-5
3.6 Power Resource Management.......................................................................................................................3-5
3.7 Code Resource Management.........................................................................................................................3-6
3.7.1 HS-SCCH Code Resource Management..............................................................................................3-6
3.7.2 HS-PDSCH Code Resource Management ...........................................................................................3-7
3.7.3 RNC-Controlled Static Code Allocation..............................................................................................3-7
3.7.4 RNC-Controlled Dynamic Code Allocation ........................................................................................3-7
3.7.5 NodeB-Controlled Dynamic Code Allocation .....................................................................................3-9
3.7.6 Dynamic Code Tree Reshuffling........................................................................................................3-10
4 User Plane....................................................................................................................................4-1
4.1 Flow Control and Congestion Control ..........................................................................................................4-1

Contents
RAN
HSDPA
iv Huawei Proprietary and Confidential
Issue 02 (2009-06-30)
4.1.1 Flow Control........................................................................................................................................4-2
4.1.2 Congestion Control..............................................................................................................................4-3
4.2 RLC and MAC-d...........................................................................................................................................4-3
4.2.1 RLC......................................................................................................................................................4-3
4.2.2 MAC-d.................................................................................................................................................4-4
4.3 MAC-hs Scheduling......................................................................................................................................4-4
4.3.1 Determining the Candidate Set ............................................................................................................4-4
4.3.2 Calculating Priorities ...........................................................................................................................4-5
4.3.3 Comparison of Four Algorithms ..........................................................................................................4-8
4.4 HARQ ...........................................................................................................................................................4-9
4.4.1 HARQ Retransmission Principles........................................................................................................4-9
4.4.2 Soft Combining During HARQ .........................................................................................................4-10
4.4.3 Preamble and Postamble....................................................................................................................4-10
4.5 TFRC Selection...........................................................................................................................................4-11
4.5.1 Basic Procedure of TFRC Selection...................................................................................................4-11
4.5.2 Determining the TBSmax.....................................................................................................................4-11
4.5.3 Determining the TBSused, Modulation Scheme, Power, and Codes....................................................4-13
4.5.4 Determining the Number of MAC-d PDUs .......................................................................................4-14
5 QoS and Diff-Serv Management ............................................................................................5-1
5.1 QoS Management..........................................................................................................................................5-1
5.2 Diff-Serv Management..................................................................................................................................5-3
5.2.1 SPI Weight Description........................................................................................................................5-3
5.2.2 Differentiated Services Based on Service Types..................................................................................5-4
5.2.3 Differentiated Services Based on User Priorities.................................................................................5-4
5.3 QoS Parameter Mapping and Configuration.................................................................................................5-5
6 Parameters ...................................................................................................................................6-1
7 Counters.......................................................................................................................................7-1
8 Glossary .......................................................................................................................................8-1
9 Reference Documents ...............................................................................................................9-1

RAN
HSDPA 1 Introduction to This Document
1-1
1 Introduction to This Document
1.1 Scope
the HSDPA functional area. It provides an overview of the main
tails regarding HSDPA control and user plane functions.
1.2 Intended Audience
s asics and have a
ork
This document is intended for:
System operators who need a general understanding of HSDPA
rking on Huawei products or systems
1.3 Change
is ument versions.
Ther ined as follows:
Feature change: refers to the change in the HSDPA feature.
orial change: refers to the change in wording or the addition of the information that
ersion.
Document Issues
as follows:
01 (2009-03-30)
Draft (2009-03-10)
This document describes
functions and goes into de
It i assumed that users of this document are familiar with WCDMA b
w ing knowledge of 3G telecommunication.
Personnel wo
History
Th section provides information on the changes in different doc
e are two types of changes, which are def
Edit
was not described in the earlier v
The document issues are
02 (2009-06-30)
Draft (2009-01-15)

RAN
HSDPA
1-2 Huawei Proprietary and Confidential
Issue 02 (2009-06-30)
02 (2009-06-30
This is the document for the second commercial release of RAN11.0.
C ( , this issue incorporate in
t e.
)
ompared with 01 2009-03-30) of RAN11.0 s the changes described
he following tabl
Change Type Change Description Parameter Change
Feature change None. None.
The description of MAC-hs Scheduling is
optimized. For details, see section 4.3
MAC-hs Scheduling. HarqRt
arqRt
ters
128KRSCLMT
The deleted parameters
are as follows:
MaxDchVoip
MaxDchAmrH
The added parame
are as follows:
8KRSCLMT
16KRSCLMT
32KRSCLMT
64KRSCLMT
256KRSCLMT
384KRSCLMT
Editorial change
Management is optimized. For details, see
section 5.2 Diff-Serv Management and 5.3
QoS Parameter Mapping and Configuration.
ters
y
3Priority
The description of QoS and Diff-Serv The added parame
are as follows:
SingalUlMBR
SingalDlMBR
StreamUlMBR
StreamDlMBR
ConverUlMBR
ConverDlMBR
ARP1Priority
ARP2Priority
ARP3Priority
ARP4Priority
ARP5Priority
ARP6Priority
ARP7Priority
ARP8Priority
ARP9Priority
ARP10Priority
ARP11Priority
ARP12Priorit
ARP1

RAN
HSDPA 1 Introduction to This Document
1-3
ARP14Priority
TrafficClass
PRIORITY
TOR
HappyBR
THPClass
THP
USER
UlGBR
DlGBR
SPI
FAC
The structure of the document is adjusted. None.
01 (2009-03-30
T ent fo lease of RAN11.0.
C aft (2 -10), this issue incorporates the following es:
)
his is the docum r the first commercial re
ompared with dr 009-03 chang
Feature change None None
Editorial change The structure of the docuement is adjusted. None
Draft (2009-03
he second draft of the document for RAN11.0.
optimizes the description.
Draft (2009-01
This ini nt for RAN11.0.
C d with issue 03 (2008-11-30) of RAN10.0, draft (2009-01-15) incorporates the
following changes:
-10)
This is t
Compared with draft (2009-01-15), draft (2009-03-10)
-15)
is the
ompare
tial draft of the docume
Change
Type
Change Description Parameter Change
The description of dynamic co
reshuffling is added in section
de tree
rNumThd
itch
3.7.6
"Dynamic Code Tree Reshuffling."
The added parameters are as follows:
CodeAdjForHsdpaUse
CodeAdjForHsdpaSw
Feature
change
cription of setting the
maximum number of retransmissions
ows:The des The added parameters are as foll

RAN
HSDPA
Issue 02 (2009-06-30)
Change Change Description Parameter Change
Type
on a service basis is added to section
MaxDchAmrHarqRt
MaxNonConverHarqRt
4.3.1 "Determining the Candidate
Set."
MaxDchVoipHarqRt
The description of HBR-based
resource allocation is added to section
4.3.2 "Calculating Priorities."
The added parameter is HappyBR.
The description of a new resource
tion
d,
r, and
The parameter RscAllocM is added
new value PowerCode_Bal.allocation method is added to sec
4.5.3 "Determining the TBSuse
Modulation Scheme, Powe
Codes."
with a
The description of HSDPA is
rewritten for readability.
NoneEditorial
change
All the parameter names are replaced
with the corresponding parameter IDs.
None

RAN
HSDPA 2 Overview of HSDPA
2-1
2 Overview of HSDPA
2.1 Genera
services on the mobile network, 3GPP Release
5 introduced HSDPA in 2005. HSDPA improves the downlink capacity, increases the user data
r ces
The characteristics of H
l Principles of HSDPA
To meet the rapidly growing demands for data
ate greatly, and redu the transmission delay on the WCDMA network.
SDPA are as follows:

2 Overview of HSDPA
RAN
HSDPA
Issue 02 (2009-06-30)
Fast scheduling Fast scheduling introduced into the NodeB determines the UEs for
data transmission in each TTI (2 ms) and dynamically allocates
resources to these UEs. It improves the usage of system resources
and increases the system capacity.
For details about how Huawei RAN implements fast scheduling, see
section 4.3 "MAC-hs Scheduling."
Fast HARQ Fast hybrid automatic repeat request (HARQ) is used to rapidly
Specifically, when the UE detects an erroneous data transmission, it
ed data and requests the NodeB to retransmit the
original data at the physical layer. Before decoding, the UE
performs soft combining of the saved data and the retransmitted
data. The combining makes full use of the data transmitted each
time and thus increases the decoding success rate. In addition, the
retransmission delay at the physical layer is reduced greatly,
compared with that at the RLC layer.
For details about how Huawei RAN implements fast HARQ, see
section 4.4 "HARQ."
request the retransmission of erroneously received data.
saves the receiv
Fast AMC To compensate for channel variations, the DCH performs power
control. To achieve this goal, HSDPA also performs fast adaptive
modulation and coding (AMC), that is, adjusts the modulation
scheme and coding rate in each TTI. AMC is based on channel
quality indicator (CQI) reported by the UE, and its purpose is to
select an appropriate transmission rate so as to meet channel
conditions. When the channel conditions are good, 16QAM can be
ide higher transmission rates. When the channel
re poor, QPSK can be used to ensure the transmission
wei RAN implements fast AMC, see
the
used to prov
conditions a
quality.
For details about how Hua
section 4.5 "TFRC Selection."
The MAC-hs, a new MAC sublayer, is introduced into the UE and NodeB to support HSDPA.
2.2 HSDPA Channels
ows the physical channels of HSDPA in the shaded area.
To support the HSDPA technologies, 3GPP defines one transport channel (HS-DSCH) and
three physical channels (HS-PDSCH, HS-SCCH, and HS-DPCCH).
Figure 2-1 sh

RAN
2-3
Figure 2-1 Physical channels of HSDPA
2.2.1 HS-DSCH and HS-PDSCH
red channel. Its TTI is fixed to 2 ms. It may be
The use of 2 ms TTI reduces the round trip time (RTT) on the Uu interface and, together with
AMC, improves the tracking of channel variations. In addition, the use of 2 ms TTI enables
and thus improves the usage of transmission
ich the HS-DSCH maps. More
2.2.2 HS-SC
HS-SCCH is a high speed shared control channel. It carries the control information related to
n,
of
imultaneously in
each TTI.
h HSDPA
UE to report
HS-DSCH is a high speed downlink sha
mapped onto one or more HS-PDSCHs.
HS-PDSCH is a high speed physical downlink shared channel. Its spreading factor is fixed to
16. According to 3GPP TS 25.433, a maximum of 15 HS-PDSCHs can be used for
transmission at the same time. The number of HS-PDSCHs per cell is configurable.
Generally, the NodeB can use the HS-PDSCH codes only allocated by the RNC. The
NodeB-controlled dynamic code allocation, however, allows the NodeB to temporarily
allocate idle codes to the HS-PDSCH. "Dynamic Code Allocation Based on NodeB" is an
optional feature.
fast scheduling and resource allocation
resources.
In each TTI, HSDPA assigns the HS-PDSCHs onto wh
HS-PDSCHs can provide higher transmission rates.
Unlike the DCH, the HS-DSCH cannot support soft handover. The reason is that this type of
handover requires different cells to use the same radio resource for sending the same data to
the UE, but the scheduling function can be performed only within the cell.
CH
the HS-DSCH. The control information includes the UE identity, HARQ-related informatio
and information about transport format and resource combination (TFRC). For each
transmission of the HS-DSCH, one HS-SCCH is required to carry the related control
information. One cell can be configured with a maximum of four HS-SCCHs. The number
HS-SCCHs determines the maximum number of UEs that can be scheduled s
2.2.3 HS-DPCCH
HS-DPCCH is a high speed dedicated physical control channel. In the uplink, eac
UE must be configured with an HS-DPCCH. This channel is mainly used by the

2 Overview of HSDPA
RAN
HSDPA
Issue 02 (2009-06-30)
the CQI and whether a transport block is correc
transport block is used for fast retransmission a
tly received. The information about the
t the physical layer. The CQI is used for AMC
2.2.4 DPCCH and DPCH/F-DPCH
control channel in the uplink. DPCH is a dedicated physical
channel in the downlink. F-DPCH is a fractional dedicated physical channel in the downlink.
edicated physical control channels in both the
case,
cell
e
MC,
hus, it reduces both unnecessary delays and processing complexity caused by Iub message
hange.
the physical layer for data
SDPA, 3GPP defines 12 UE categories. These UEs support different
peak
supp
detai
2.4 HSDPA Functions
HSD
2.4.1 HSDPA C
The aintaining HS-DSCH connections and
na
Figu
and m
and scheduling to allocate Uu resources.
DPCCH is a dedicated physical
The HSDPA UE must be configured with d
uplink and the downlink. The uplink DPCCH is used for providing reference information
about the transmit power of HSDPA channels. In addition, it is used for closed-loop power
control by working with the DPCH or F-DPCH. In SRB over HSDPA mode, the downlink
channel can be established on the F-DPCH without the dedicated assisted DPCH. In this
a maximum of 10 UEs use an SF256 to transmit the TPC, thus saving a large amount of
downlink codes.
2.3 Impact of HSDPA on NEs
HSDPA has an impact on the RNC, NodeB, and UE.
On the control plane of the network side, the RNC processes the signaling about HSDPA
configuration, HS-DSCH related channel configuration, and mobility management. On the
user plane of the network side, the RLC layer and MAC-d of the RNC are unchanged. At th
NodeB, the MAC-hs is added to implement HSDPA scheduling, Uu resource allocation, A
and Iub flow control. The MAC-hs implements these management functions in a short time.
T
exc
On the UE side, the MAC-hs is added between the MAC-d and
reception. To support H
rates at the physical layer, ranging from 912 kbit/s to 14 Mbit/s. The UE of category 10
orts the highest rate. The UE of category 11 or 12 supports only the QPSK mode. For
ls, see 3GPP TS 25.306. Huawei RAN supports all the UE categories.
PA functions are implemented on the HSDPA control plane and user plane.
ontrol Plane Functions
control plane is responsible for setting up and m
ma ging cell resources.
re 2-2 shows the HSDPA control plane functions based on the service connection setup
aintenance procedure.

RAN
2-5
Figure 2-2 HSDPA control plane functions
Bearer mapping
etwork side to configure the RAB during the setup
of a service connection in the cell. The network side then configures bearer channels for
es of
ient,
to
of the
For details, see section 3.3 "Mobility Management."
ad
"Channel Switching."
The HSDPA control plane functions are described as follows:
The bearer mapping is used by the n
the UE based on the requested service type, service rate, UE capability, and cell
capability.
For details, see section 3.1 "Bearer Mapping."
Access control
Access control, a sub-function of load control, checks whether the current resourc
the cell are sufficient for the service connection setup. If the resources are insuffic
intelligent access control is triggered. If the resources are sufficient, the service
connection can be set up.
For details, see section 3.2 "Access Control."
Mobility management
For the established HS-DSCH connection, mobility management decides whether
switch it to another cell for providing better services, based on the channel quality
UE.
Channel switching
Channel switching is responsible for switching the transport channel among the
HS-DSCH, DCH, and FACH based on the requirements of mobility management or lo
control.
For details, see section 3.4
Load control
When the cell load increases, the load control function adjusts the resources configured
for the established radio connections to avoid cell overload.
For details, see section 3.5 "Load Control."
Resource management

2 Overview of HSDPA
RAN
HSDPA
Issue 02 (2009-06-30)
Resource management coordinates the power resource between the HS-DSCH and the
DCH and the code resource between the HS-SCCH and the HS-PDSCH. The downlink
power and codes are the bottleneck resources of the cell. Resource management can
increase the HSDPA capacity.
Power resource management reserves power for channels of different types and allocates
power for them. For details, see section 3.6 "Power Resource Management."
Code resource management allocates and reserves code resources for channels of
different types. In addition, it collects and reshuffles idle code resources.
For details, see section 3.7 "Code Resource Management."
2.4.2 HSDPA User Plane Functions
After the service is set up, the user plane is responsible for implementing data transmission.
Figure 2-3 shows the HSDPA user plane functions based on the data processing procedure.
Figure 2-3 HSDPA user plane functions
The service data carried on the HS-DSCH is passed to the RLC layer and MAC-d of the RNC
for processing and encapsulation. Then, the MAC-d PDU is formed and passed through the
Iub/Iur interface to the NodeB/RNC. To avoid congestion, the flow control and congestion
control functions control the traffic on the Iub/Iur interface through the HS-DSCH frame
protocol (3GPP TS 25.435).
After the MAC-d PDU is received by the NodeB, it is passed through the MAC-hs to the
physical layer and then sent out through the Uu interface. The MAC-hs provides MAC-hs
scheduling, TFRC selection, and HARQ. MAC-hs scheduling determines the HSDPA users in
the cell for data transmission. TFRC selection determines the transmission rates and Uu
resources to be allocated to the HSDPA UEs. HARQ is used to implement the hybrid
automatic repeat request function.

RAN
HSDPA 3 Control Plane
3-1
3 Control Plane
This chapter consists of the following sections:
l
Load Control
Power Resource Management
Management
3.1 Bearer Mapping
ervices of multiple types and service combinations, as listed in
T -1.
Table 3-1 Bearer m
Bearer Mapping
Access Contro
Mobility Management
Channel Switching
Code Resource
The HS-DSCH can carry s
able 3
apping
CN Service Type Can Be Carried on Optional Feature?
Domain HS-DSCH?
- Signaling (SRB) Yes
RB over HSDPA
Yes
Feature name: S
Voice Yes Yes
Feat
HSP
ure name: CS Voice over
A/HSPA+
Videophone No No
CS
Streaming No No
PS al YesConversation Yes
Feature name: VoIP over
HSPA/HSPA+

3 Control Plane
RAN
HSDPA
Issue 02 (2009-06-30)
CN Service Type Can Be Carried on Optional Feature?
Domain HS-DSCH?
Streaming Yes Yes
Feature name: Streaming Traffic
Class on HSDPA
Interactive Yes No
Background Yes No
IMS signaling Yes
e: IMS Signaling
over HSPA
Yes
Feature nam
MBMS PTP Yes Yes
Feature name: MBMS P2P ove
HSDPA
r
During the service setup, the RNC selects appropriate channels based on the UE capability,
cell capability, and service parameters to optimize the use of cell resources and ensure the
QoS. Huawei RAN supports the setting of the types of RABs carried on the HS-DSCH
ice requirements. For details, see the Radio Bearers Parameter Description.
bearer management of HSDPA over Iur. "HSDPA over Iur" is an optional
3.2 Access
ection can be set up under the
ntation of this
function requires the support of channel switching.
A UE to access an inter-frequency neighboring cell that has
urce cell. The purpose is to achieve load balance between the
xperience. This is HSDPA directed retry decision (DRD), an
3.3 Mobili
The DCH supports soft handover, and therefore downlink data can be concurrently sent out
according to serv
Huawei supports
feature.
Control
Access control determines whether an HS-DSCH conn
precondition that the QoS is ensured. The determination is based on the status of cell
resources and the situation of Iub/Iur congestion. When the resources are insufficient, the
HS-DSCH is switched to the DCH and only the DCH connection is set up. When the
resources are sufficient, the DCH is switched to the HS-DSCH. The impleme
Access control allows the HSDP
the same coverage area as the so
cells and improve HSDPA user e
optional feature. For details, see the Load Control Parameter Description.
ty Management
from all the cells in the active set in DCH transmission. In comparison, the HS-DSCH does
not support soft handover, and therefore downlink data can be sent out only from the
HS-DSCH serving cell and inter-cell handover has to be performed through the change of the
serving cell. Thus, HSDPA mobility management focuses on the change of the HS-DSCH
serving cell.

RAN
3-3
For the UE with the HS-DSCH service, the best cell in the active set acts as the HS-DSCH
serving cell. When the best cell changes, the UE disconnects the HS-DSCH from the source
cell and attempts to set up a new HS-DSCH connection with the new best cell. For details,
the Handover Parameter Description. By changing the HS-DSCH switch
see
ing threshold, you
-DSCH connection with the target cell, the
obility management, switches the HS-DSCH to
available, the channel switching function
switches the DCH back to the HS-DSCH. When the HSDPA user returns from the DCH cell
to the HSDPA cell, the DCH is set up to ensure successful handover. A certain period later
annel switching function switches the DCH to the HS-DSCH. For
hannel Switching."
3.4 Channel Switching
ed, the UE can stay in a new state, CELL_DCH (with
and
can modify the conditions for triggering the change of the best cell. Lowering this threshold
can increase both the handover frequency and the sensitivity of HS-DSCH switching to signal
variations in the serving cell. Raising this threshold can reduce the handover frequency but
may increase the probability of the HS-DSCH service being discontinuous or even dropping
on the cell edge. For the HS-DSCH service, Huawei supports inter-cell intra-frequency
handover, inter-cell inter-frequency handover, and inter-RAT handover.
Mobility management may trigger the switching from the HS-DSCH to the DCH. If the UE
with the HS-DSCH service cannot set up the HS
channel switching function, together with m
the DCH. When the HS-DSCH connection is
after the handover, the ch
details, see section 3.4 "C
"HSDPA over Iur" is an optional feature.
After the HS-DSCH is introduc
HS-DSCH). Thus, there are additional transitions between CELL_DCH (with HS-DSCH)
CELL_FACH and transitions between CELL_DCH (with HS-DSCH) and CELL_DCH even
when both the cell and the UE support the HS-DSCH, as shown in Figure 3-1.
Figure 3-1 UE state transition
T i
Table 3-2 New state transition and new channel switching
able 3-2 lists new state transition and new channel sw tching.
New State Transition New Channel Switching
CELL_DCH (with HS-DSCH) <-> CELL_FACH HS-DSCH <-> FACH
CELL_DCH (with HS-DSCH) <-> CELL_DCH HS-DSCH <-> DCH

3 Control Plane
RAN
HSDPA
Issue 02 (2009-06-30)
Here, the switching between HS-DSCH and FACH can be triggered by traffic volume, whic
is similar to the switching between
h
DCH and FACH.
t some cells support the HS-DSCH but
p, the channel switching
ability and UE capability
oving, the
to ensure
When the cell load is too high, load control may also trigger the switching from the
HS-DSCH to the FACH to relieve congestion. For details, see the Load Control Parameter
Description. When the cell load becomes low, channel switching aids load control in
attempting to switch the transport channel back to the HS-DSCH. For details, see the Rate
Control Parameter Description.
As the HS-DSCH is introduced later, it is inevitable tha
others do not. This is also the case with UEs. When a service is set u
function selects an appropriate bearer channel based on the cell cap
to ensure the QoS while efficiently using the cell resources. When the user is m
channel switching function adjusts the channel type based on the UE capability
service continuity while improving user experience.
Figure 3-2 Relations between channel switching and other functions
Trigg
ser
f
Th
con a case, the possible scenarios are as follows:
to a
is case, the DCH connection is also set up because
In on eviously, the DCH connection is set up in a cell supporting the
e neighboring cell supporting the HS-DSCH.
Then
ers for switching from the HS-DSCH to the DCH are as follows:
The HS-DSCH is selected during the service setup but neither the resources of the
ving cell nor the resources of the inter-frequency same-coverage neighboring cell are
su ficient. In such a case, the HS-DSCH is switched to the DCH.
e HS-DSCH serving cell changes. The UE attempts to set up a new HS-DSCH
nection with the new best cell. In such
− If the new best cell does not support the HS-DSCH, the UE cannot set up the
HS-DSCH connection. In this case, the HS-DSCH is switched to the DCH.
− If the new best cell supports the HS-DSCH but a new HS-DSCH connection cannot
be set up because the resources are insufficient, the DCH connection is set up and the
HS-DSCH is switched to this DCH.
The user moves from a cell supporting the DCH but not supporting the HS-DSCH
cell supporting the HS-DSCH. In th
the DCH supports soft handover, which can increase the inter-cell handover success rate.
e of the cases described pr
HS-DSCH or in an inter-frequency same-coverag
, the DCH is switched to the HS-DSCH by either of the following mechanisms:

RAN
3-5
ically attempts to switch the
DCH to the HS-DSCH.
Channel switching based on traffic volume
e traffic volume of the UE increases and the RNC receives an event 4A report,
hanism attempts to switch the DCH to the HS-DSCH. For details on the event
3.5 Load C
), and
bearer type. When the cell load is high, the basic congestion control selects some HSDPA
users for handover to an inter-frequency same-coverage neighboring cell or an inter-RAT
ll load is too high, the overload congestion
he switching to a common channel or releases
3.6 Power
ent determines the transmit power of the HS-PDSCH, HS-SCCH,
Gene
re
c
1. sources are first reserved for common physical channels and
2. k control channels,
rces are
ownlink control channel
HS-SCCH. For details, see the Power Control Parameter Description. The remaining
power resources are allocated to the traffic channel HS-PDSCH.
For details on power resource allocation, see section 4.5 "TFRC Selection."
Figure 3-3 shows the dynamic HSDPA power resource allocation.
Channel switching based on timer
After the DCH connection is set up, this mechanism period
When th
this mec
4A report, see the Rate Control Parameter Description.
ontrol
When the cell is congested, load control selects some users (including HSDPA users) for
congestion relief. The selection is based on the integrated priority, which considers the
allocation retention priority (ARP), traffic class (TC), traffic handling priority (THP
neighboring cell with lower load. When the ce
control selects some HSDPA BE services for t
some HSDPA services. For details, see the Load Control Parameter Description.
Resource Management
Power resource managem
and HS-DPCCH.
rally, an HSDPA cell has the same coverage as the corresponding R99 cell. To improve
the source usage in this case, the downlink power resources of HSDPA can be dynamically
allo ated as follows:
The downlink power re
allocated to the DPCH. The remaining power resources are available for HSPA,
including HSUPA and HSDPA.
The HSPA power resources are first allocated to the HSUPA downlin
including the E-AGCH, E-RGCH, and E-HICH. The remaining power resou
available for HSDPA.
3. The HSDPA power resources are first allocated to the d

3 Control Plane
RAN
HSDPA
Issue 02 (2009-06-30)
Figure 3-3 Dynamic HSDPA power resource allocation
Every TTI, the NodeB detects the power usage of R99 channels to determine the power
available for HSPA. To reserve the power for R99 power control itself, the power margin
PwrMgn needs to be set on the NodeB side. In addition, the power allocated to HSPA must
not exceed the maximum permissible power HspaPower, which can be set on the RNC side.
For details on uplink HS-DPCCH power control, see the Power Control Parameter
Description.
"HSDPA over Iur" is an optional feature.
3.7 Code Resource Management
Code resource management allocates code resources to the HS-SCCH and HS-PDSCH.
The NodeB supports HS-DSCH transmissions to multiple users in parallel in a TTI. If more
than one HS-PDSCH code can be allocated by the NodeB, then code multiplexing can be used
to allocate the codes to multiple users so as to improve resource usage and system throughput.
"Time and HS-PDSCH Code Multiplexing" is an optional feature.
3.7.1 HS-SCCH Code Resource Management
Each HS-SCCH uses an SF128 code. The number of HS-SCCHs determines the maximum
number of HSDPA users that can be scheduled simultaneously in a TTI. Generally, the
number of HS-SCCHs depends on the traffic characteristics of the cell. The default number is
4, which is specified by the parameter HsScchCodeNum on the RNC side. If the default
setting is used, the HS-PDSCH can use only 14 SF16 codes. To enable the HS-PDSCH to use
15 SF16 codes, you are advised to configure 2 HS-SCCHs.

RAN
3-7
3.7.2 HS-PDSCH Code Resource Management
The DPCH and the HS-PDSCH coexist in a cell. Therefore, sharing the cell code resources
between them to improve the resource usage is of critical importance in HSDPA code resource
management.
Huawei supports both RNC-level and NodeB-level code resource management.
RNC-controlled static or dynamic code allocation is enabled through the parameter
AllocCodeMode. NodeB-controlled dynamic code allocation is enabled through the
parameter DynCodeSw. For details, see the following sections.
The dynamic code allocation controlled by the NodeB is more flexible than that controlled by
the RNC. It shortens the response time and saves the Iub signaling used for code reallocation.
Huawei recommends the following code allocation modes, where the first mode is preferred:
Configure the RNC to use static code allocation and the NodeB to use dynamic code
allocation.
If the NodeB does not support dynamic code allocation, configure the RNC to use
dynamic code allocation.
If not all the NodeBs controlled by an RNC support dynamic code allocation, the
RNC-controlled dynamic code allocation is recommended. In this case, the NodeB-controlled
dynamic code allocation can also be enabled for those supporting NodeBs.
3.7.3 RNC-Controlled Static Code Allocation
If the RNC-controlled static code allocation is used, the number of reserved HS-PDSCH
codes is specified by the parameter HsPdschCodeNum on the RNC side. Based on the
number, the RNC reserves codes for the HS-PDSCH. The DPCH, HS-SCCH, and common
channels use the other codes. The parameter HsPdschCodeNum can be set on the basis of the
traffic characteristics of the cell. If there are more HSDPA users and the traffic is high, the
parameter value can be increased. If there are more DCH users and the HSDPA traffic is low,
the parameter value can be decreased. A maximum of 15 codes can be allocated to the
HS-PDSCH.
Figure 3-4 shows the RNC-controlled static code allocation.
Figure 3-4 RNC-controlled static code allocation
3.7.4 RNC-Controlled Dynamic Code Allocation
If the RNC-controlled dynamic code allocation is used, the minimum number of available
HS-PDSCH codes is specified by the parameter HsPdschMinCodeNum on the RNC side.
The purpose of this setting is to prevent too many DCH users from being admitted and to
ensure the basic data transmission of the HS-PDSCH. In addition, the maximum number of
available HS-PDSCH codes is specified by the parameter HsPdschMaxCodeNum. The

3 Control Plane
RAN
HSDPA
Issue 02 (2009-06-30)
purpose of this setting is to prevent too many codes from being allocated for the HS-PDSCH
and to prevent DCH users from preempting codes during admission.
The number of codes that can be shared between HS-PDSCH and DPCH is equal to the value
of HsPdschMaxCodeNum minus the value of HsPdschMinCodeNum, as shown in Figure
3-5. When a code that can be shared is idle, it can be allocated to the HS-PDSCH if the idle
code is adjacent to the allocated HS-PDSCH codes.
Figure 3-5 RNC-controlled dynamic code allocation
Adding an HS-PDSCH Code
Figure 3-6 shows how to add an HS-PDSCH code. The solid dots represent the allocated
codes, and the circles represent the idle codes.
Figure 3-6 Adding an HS-PDSCH code
After a DCH RL is released or reconfigured (for example, because the spreading factor
becomes larger), the RNC adds an HS-PDSCH code if the following conditions are met:
The code adjacent to the allocated HS-PDSCH codes is idle.
After the code is added, the minimum spreading factor of the remaining codes is smaller
than or equal to the value of CellLdrSfResThd.
The parameter CellLdrSfResThd set on the RNC side is used to reserve codes for new users,
to avoid congestion due to code insufficiency, and to avoid unnecessary reshuffling of the
code tree.

RAN
3-9
Releasing an HS-PDSCH Codes
Figure 3-7 shows how to release an HS-PDSCH code. The solid dots represent the allocated
codes, and the circles represent the idle codes.
Figure 3-7 Releasing an HS-PDSCH code
If idle DPCH codes are insufficient when a DCH RL is set up, added, or reconfigured (for
example, because the spreading factor becomes smaller), the RNC preempts HS-PDSCH
codes in the shared codes for the DPCH. In addition, if the minimum spreading factor of idle
DPCH codes is greater than the value of CellLdrSfResThd, the RNC can also reallocate
some HS-PDSCH codes to the DPCH. The reallocated code number must be the smallest one
of the available shared codes.
3.7.5 NodeB-Controlled Dynamic Code Allocation
Generally, the NodeB can use the HS-PDSCH codes only allocated by the RNC. The
NodeB-controlled dynamic code allocation, however, allows the NodeB to temporarily
allocate idle codes to the HS-PDSCH.
Figure 3-8 NodeB-controlled dynamic code allocation
Every TTI, the NodeB detects the SF16 codes that are not allocated to the HS-PDSCH. If
such an SF16 code or any of its subcodes is allocated by the RNC to the DCH or a common
channel, this SF16 code is regarded as occupied. Otherwise, it is regarded as unoccupied.
Therefore, the available HS-PDSCH codes include the codes reserved by the RNC and the
idle codes adjacent to the allocated HS-PDSCH codes. Every time the RNC allocates or
release HS-PDSCH codes, it notifies the NodeB through Iub signaling and the NodeB
performs the corresponding processes.
For example, the RNC reserves the SF16 codes numbered 11 to 15 for the HS-PDSCH and
those numbered 0 to 5 for the DCH and common channels in a TTI. Thus, the HS-PDSCH can
use the codes numbered 6 to 15 in this TTI.

3 Control Plane
RAN
HSDPA
Issue 02 (2009-06-30)
If the setup of an RL requires a DPCH code that is already allocated by the NodeB to the
HS-PDSCH, the NodeB releases this code and sends an NBAP message to the RNC,
indicating that the RL is set up successfully. Then, the DCH uses this code. After the DCH
releases it, the HS-PDSCH can use this code again.
"Dynamic Code Allocation Based on NodeB" is an optional feature.
3.7.6 Dynamic Code Tree Reshuffling
Regardless of whether dynamic code allocation is controlled by the RNC or the NodeB, the
number of continuous codes available for the HS-PDSCH shall be maximized. The dynamic
code tree reshuffling function can achieve this goal by reallocating DPCH codes.
When the minimum spreading factor of the remaining idle codes in a cell is greater than the
value of CellLdrSfResThd, the RNC reshuffles the codes used by the DPCH to provide more
continuous SF16 codes for HSDPA. This function can be enabled or disabled by the
parameter CodeAdjForHsdpaSwitch on the RNC side.
In addition, the threshold number of users that can be reshuffled needs to be specified by the
parameter CodeAdjForHsdpaUserNumThd. If the number of users on a subtree is smaller
than or equal to this parameter value, this subtree can be reshuffled. Otherwise, it cannot be
reshuffled. This parameter limits the number of users that can be reshuffled each time, to
prevent too many users from being reshuffled in a short time and thus to avoid affecting user
experience.
Figure 3-9 Dynamic code tree reshuffling

RAN
HSDPA 4 User Plane
4-1
4 User Plane
ntrol and Congestion Control
MAC-hs Scheduling
HARQ
4.1 Flow C
, or
used to
deB. HSDPA
3GPP TS
25.435). They are implemented for each MAC-hs queue through the Capacity Request
message sent by the RNC and the Capacity Allocation message sent by the NodeB.
Figure 4-1 shows the basic principles of flow control and congestion control.
Flow Co
RLC and MAC-d
TFRC Selection
ontrol and Congestion Control
HSDPA flow control and congestion control are used to control the HSDPA data flow on the
Iub and Iur interfaces. HSDPA data packets are sent through the Iub interface to the NodeB
and then through the Uu interface to the UE. Thus, congestion may occur on the Uu, Iub
Iur interface. Flow control is used to relieve Uu congestion, and congestion control is
relieve Iub/Iur congestion. The two types of control are implemented by the No
flow control and congestion control are part of the HSDPA Iub frame protocol (

4 User Plane
RAN
HSDPA
Issue 02 (2009-06-30)
Figure 4-1 Basic principles of Iub flow control and congestion control
4.1.1 Flow Control
For each MAC-hs queue, flow control calculates the pre-allocated Iub bandwidth based on the
Uu transmission rate and the amount of data buffered in the NodeB. The Uu transmission rate
of the MAC-hs queue is determined by the scheduling algorithm. For each MAC-hs queue, if
the Iub transmission rate is higher than the Uu transmission rate, the data packets are buffered.
Too much data buffered in the NodeB leads to transmission delay and even packet loss.
Therefore, each MAC-hs queue should not have too much data buffered in the NodeB. On the
other hand, it should keep a certain amount of data to avoid wasting the Uu resources due to
no data to transmit.
The flow control procedure is as follows:
1. The NodeB measures the buffered data amount of each MAC-hs queue and the average
Uu transmission rate.
2. The NodeB estimates the buffering time based on the measurements.
3. The NodeB adjusts the Iub bandwidth pre-allocated to the MAC-hs queue.
The pre-allocated Iub bandwidth is adjusted as follows:
If the buffering time is too short, you can infer that the RNC slows down the data
transmission, that is, the Iub transmission rate is lower than the Uu transmission rate. In
such a case, the pre-allocated Iub bandwidth is adjusted to a value greater than the
average Uu transmission rate.
If the buffering time is appropriate, the pre-allocated Iub bandwidth is adjusted to the
average Uu transmission rate.
If the buffering time is too long, the pre-allocated Iub bandwidth is adjusted to a value
smaller than the average Uu transmission rate.

RAN
HSDPA 4 User Plane
4-3
4.1.2 Congestion Control
The Iub bandwidth may be lower than the Uu bandwidth. If the RNC uses the Iub bandwidth
pre-allocated to each MAC-hs queue, the Iub bandwidth for HSDPA is insufficient. This may
lead to congestion and even packet loss.
The amount of data to be transmitted is sent by the RNC to each MAC-hs queue through the
Capacity Request message. Based on this amount and the total Iub bandwidth available for
HSDPA, the congestion control function adjusts the bandwidth pre-allocated to each MAC-hs
queue. Thus, congestion control ensures that the total bandwidth actually allocated to all the
MAC-hs queues is not higher than the total available Iub bandwidth.
The total Iub bandwidth available for HSDPA depends on the variations in HSDPA packet
delay and the situation of packet loss. HSDPA shares the bandwidth with the DCH and control
signaling, and the DCH and control signaling has higher priorities than HSDPA. Thus, when
the HSDPA packet delay or packet loss increases, you can infer that the number of DCHs or
the amount of control signaling increases. In such a case, the bandwidth available for HSDPA
decreases and the bandwidth actually allocated for HSDPA decreases.
For details on congestion control, see the Transmission Resource Management Parameter
Description.
For the Iur interface, flow control and congestion control are also applied. The control principles and
processing procedures are the same as those for the Iub interface.
4.2 RLC and MAC-d
4.2.1 RLC
One of the main purposes of HSDPA is to reduce latency by handling retransmissions at
NodeB level. Retransmissions, however, may still be triggered at the RLC layer of the RNC
under the following circumstances:
The NodeB misinterprets an NACK sent by the UE.
The number of HARQ retransmissions exceeds the maximum permissible number.
The data buffered in the NodeB is lost when the HS-DSCH serving cell changes.
Therefore, HARQ retransmission cannot totally replace RLC retransmission, which is
described in 3GPP TS 25.322. For services with high requirements for data transmission
reliability, Huawei recommends that the RLC acknowledged mode (AM) also be used to
ensure correct transmission on the Uu interface even when the services such as the BE service
are carried on HSDPA channels.
Before the introduction of HSDPA, the size of an RLC PDU is usually 336 bits, where 320
bits are for the payload and 16 bits for the RLC header. Without additional overhead, the
MAC PDU is of the same size as the RLC PDU. According to the 3GPP specifications, a
maximum of 2,047 RLC PDUs can be transmitted within an RLC window, and the RTT at the
RLC layer is about 100 ms (50 TTIs). In this condition, the maximum peak rate can only be
336 bits x (2047/50)/2 ms = 6.88 Mbit/s. To reach higher rates, an RLC PDU of 656 bits is
introduced, where 640 bits are for the payload and 16 bits for the RLC header. The RLC PDU
size can be set for each typical service. For high-speed services, the size is set to 656 bits by
default.

4 User Plane
RAN
HSDPA
Issue 02 (2009-06-30)
4.2.2 MAC-d
The MAC-d functionality is unchanged after the introduction of HSDPA. The HS-DSCH
bearers are mapped onto MAC-d flows on the Iub/Iur interface. Each MAC-d flow has its
own priority queue.
The theoretical peak rate of HSDPA on the Uu interface is 14.4 Mbit/s. It is calculated on the
assumption that the chip rate of WCDMA is 3.84 Mcps, the spreading factor for HSDPA is
SF16, the maximum number of available codes is 15, and the gain of 16QAM is 4. Thus, the
rate is 3.84 Mcps/16 x 15 x 4 = 14.4 Mbit/s.
Limited by many factors, the theoretical peak rate of 14.4 Mbit/s is unreachable in actual
situations. The UE capability is one factor. For example, 3GPP specifies that the UE of
category 10 can use a maximum of 15 codes and receive a transport block with a maximum of
27,952 bits. For details, see 3GPP TS 25.306. Thus, the theoretical peak rate is 27952 bits/2
ms = 13.976 Mbit/s.
In addition, the RLC PDU size is fixed to 656 bits, and a transport block of 27,952 bits can
contain a maximum of 42 PDUs. Thus, the maximum RLC payload rate is (656 bits – 16 bits)
x 42/2 ms = 13.44 Mbit/s.
In practice, the radio channel quality, retransmission probability, and available power also
need to be considered. Therefore, the UE of category 10 cannot reach 13.44 Mbit/s at the RLC
layer in most tests.
4.3 MAC-hs Scheduling
With the limited Uu resources for HSDPA in a cell, the user expects to maximize the service
rate while the telecom operator expects to maximize the system capacity. MAC-hs scheduling
is used to coordinate the Uu resources, user experience, and system capacity. It is
implemented at the NodeB MAC-hs.
The scheduling algorithm consists of two steps. At first, the algorithm determines which
initial transmission queues or retransmission processes can be put into the candidate set for
scheduling. Then, the algorithm calculates their priorities based on factors such as the CQI,
user fairness, and differentiated services. If the algorithm is weighted more towards the
channel quality of the UE, the HSDPA cell can have a higher capacity but user fairness and
differentiated services may be affected. If the algorithm is weighted more towards user
fairness and differentiated services, the system capacity may be affected.
Huawei provides four scheduling algorithms: maximum C/I (MAXCI), round-robin (RR),
proportional fair (PF), and Enhanced Proportional Fair (EPF). The EPF algorithm is optional.
4.3.1 Determining the Candidate Set
The candidate for scheduling contains new data packets (hereinafter referred to as initial
transmission queues) or data packets to be retransmitted (hereinafter referred to as
retransmission processes), with the following exceptions:
If the UE starts the compressed mode, its data cannot be put into the candidate set during
the GAP.
If the UE category requires the UE to wait for several TTIs before it can be scheduled
again, its data cannot be put into the candidate set in this period. The UE of category 1 or
2 needs to wait for 3 TTIs, and the UE of category 3, 4, and 11 must wait for 2 TTIs.

RAN
HSDPA 4 User Plane
4-5
If the number of retransmissions of a data packet reaches or exceeds the maximum
number, the data of this UE cannot be put into the candidate set. The data should be
discarded.
Huawei supports that the maximum number of retransmissions is set on a service basis:
− MaxNonConverHarqRt: the maximum number of non-conversational service
retransmissions in the CELL_DCH state
Other user data can be put into the candidate set.
4.3.2 Calculating Priorities
Four algorithms are available for calculating the priorities of data packets in the candidate set.
The scheduling policies vary according to the algorithms for calculating the priorities of data
packets. The algorithm to be used is specified by the parameter SM on the NodeB LMT.
MAXCI Algorithm
The retransmission processes unconditionally have higher priorities than the initial
transmission queues. The retransmission processes are sorted in first-in first-out (FIFO) mode.
The initial transmission queues are sorted in the CQI order. A higher CQI means a higher data
priority.
The MAXCI algorithm aims to maximize the system capacity but cannot ensure user fairness
and differentiated services.
RR Algorithm
transmission queues. The retransmission processes are sorted in FIFO mode. The initial
transmission queues are sorted in the order of the waiting time in the MAC-hs queue. A longer
waiting time means a higher data priority.
The RR algorithm aims to ensure user fairness but cannot provide differentiated services. Not
considering the CQI reported by the UE leads to lower system capacity.
PF Algorithm
transmission queues. The retransmission processes are sorted in FIFO mode. The initial
transmission queues are sorted in the order of R/r. Here, R represents the throughput
corresponding to the CQI reported by the UE, and r represents the throughput achieved by the
UE. A greater R/r value means a higher data priority.
The PF algorithm aims to make a tradeoff between system capacity and user fairness. It
provides the user with an average throughput that is proportional to the actual channel quality.
The system capacity provided by PF is between the system capacity provided by RR and that
provided by MAXCI.
EPF Algorithm
The EPF algorithm can meet the requirements of telecom operators related to user fairness
and differentiated services and also provide a high system capacity.

4 User Plane
RAN
HSDPA
Issue 02 (2009-06-30)
Firstly, priorities are determined on the basis of service types. The EPF algorithm
distinguishes between delay-sensitive data and throughput-sensitive data based on the QoS
requirements.
The amount of delay-sensitive data is generally small. The transmission delay of
delay-sensitive data should be as short as possible. When the transmission delay reaches a
specified threshold, data packets are discarded. The delay-sensitive data includes the
following data:
SRB signaling
VoIP and AMR service data whose waiting time approaches the value of the discard
timer
The amount of a throughput-sensitive data is generally small. A higher transmission rate
brings greater user satisfaction. The throughput-sensitive data includes the following data:
BE service data
Streaming service data
IMS data
VoIP and AMR service data whose waiting time is far from the value of the discard timer
The EPF algorithm meets the basic QoS requirements of users. For delay-sensitive data, the
transmission delay must not exceed the maximum permissible delay. For throughput-sensitive
data, the transmission rate must not be lower than the GBR. Users require higher QoS for
delay-sensitive data. Therefore, the delay-sensitive data has a higher priority than the
throughput-sensitive data.
Secondly, for delay-sensitive data or throughput-sensitive data, the EPF algorithm
distinguishes between retransmission processes and initial transmission queues. The
retransmission processes unconditionally have higher priorities than the initial transmission
queues.
Thirdly, the priorities of the initial transmission queues are calculated for delay-sensitive data
or throughput-sensitive data. The following factors are considered: the waiting time, CQI
reported by the UE, throughput achieved by the UE, guaranteed bit rate (GBR), scheduling
priority indicator (SPI) weight, happy bit rate (HBR), and power consumed in the queue for a
certain period. The impacts of these factors on the priority calculation are as follows:
For the delay-sensitive data, a longer waiting time means a higher data priority.
For the throughput-sensitive data, a greater R/r value means a higher data priority. Here,
R represents the throughput corresponding to the CQI reported by the UE, and r
represents the throughput achieved by the UE.
The UEs with the rates lower than the GBR have higher priorities than those with the
rates already reaching the GBR.
A higher SPI weight means a higher data priority.
A larger difference between the actual rate and the HBR means a higher data priority.
When the resource limitation switch (RscLmSw) is on, the algorithm allocates the
lowest priority to a queue whose power consumption exceeds the threshold. RscLmSw
is used to prevent the users in areas with poor coverage from consuming too many cell
resources so that there is no decrease in system capacity. The ratio of the maximum
available power of a queue to the total power of the cell depends on the GBR, as listed in
Table 4-1.
By calculating the priority of each queue, the scheduling algorithm achieves the following:

RAN
HSDPA 4 User Plane
4-7
When the system resources are sufficient to meet the basic QoS requirements of all users,
the transmission delay of delay-sensitive data is within the permissible range and the
transmission rate of throughput-sensitive data is not lower than the GBR. High-priority
users can obtain more resources for higher QoS.
When the system resources are insufficient to meet the basic QoS requirements of all
users, delay-sensitive data has higher priorities than throughput-sensitive data.
High-priority users can obtain more resources to ensure the basic QoS.
Fourthly, special processing is performed.
Differentiated services based on SPI weights are provided. Different services have
different service types, and different users have different user priorities. Therefore, the
scheduling function needs to consider these two factors to provide differentiated services.
SPI is a parameter specified on the basis of service types and users priorities. The
parameter SPIweight can be specified according to the SPI to provide differentiated
services. This parameter is specified on the RNC, and its value ranges from 0% to 100%.
The SPI weight affects the calculation of queue priorities. It is used to quantify the
differentiated services. If all the rates of throughput-sensitive services with different SPI
weights exceed or none of the rates exceeds their GBRs, the proportion of SPI weights
determines the proportion of rates among users. For example, for three
throughput-sensitive service users with the same channel quality, if their GBRs are not
configured and the proportion of SPI weights is 100:50:30, the proportion of actual rates
is close to 100:50:30.
Differentiated services based on SPI weights are optional.
Users with poor channel quality are prevented from consuming too many radio resources.
If a user in a poor-coverage area, for example, at the edge of a cell, has a high priority,
too many radio resources may be consumed to meet the QoS requirement. In this case,
the QoS of other users may be affected. To solve this problem, resource restriction
parameters such as 8KRSCLMT, 16KRSCLMT, 32KRSCLMT, 64KRSCLMT,
128KRSCLMT, 256KRSCLMT, and 384KRSCLMT are defined to restrict the
maximum power consumption of each user. They are configured on the NodeB
according to the GBRs.
Table 4-1 Default maximum ratios based on the GBR
GBR (kbit/s) Maximum Ratio
8 10%
16 10%
32 15%
64 15%
128 20%
256 25%
384 30%
The HBR is configured. The HBR determines the throughput expected by the user based
on a study on user experience. When the rate for a user reaches the HBR, the scheduling
probability for the user is decreased. Therefore, the scheduling probability of the users
with rates lower than the HBR is increased. In this way, more users can obtain satisfying

4 User Plane
RAN
HSDPA
Issue 02 (2009-06-30)
services. The HBR is specified by the parameter HappyBR on the RNC side. The setting
can be based on user levels, including gold, silver, and copper.
For details on the parameters related to QoS management, such as the GBR, SPI, SPI weight,
and HBR, see section 5.3 "QoS Parameter Mapping and Configuration."
The EPF algorithm is optional.
4.3.3 Comparison of Four Algorithms
Table 4-2 lists the factors considered in the four scheduling algorithms.
Table 4-2 Factors considered in the four scheduling algorithms
Factor MAXCI RR PF EPF
Service type No No No Yes
Initial transmission or
retransmission
Yes Yes Yes Yes
Maximum power No No No Yes
Waiting time No Yes No Yes
CQI Yes No Yes Yes
Actual throughput No No Yes Yes
SPI No No No Yes
GPR No No No Yes
HBR No No No Yes
Table 4-3 lists the effects of the four scheduling algorithms.
Table 4-3 Effects of the four scheduling algorithms
Item MAXCI RR PF EPF
System capacity Highest High Higher Higher
User fairness Not guaranteed Best Guaranteed Guaranteed
Differentiated
services
Not guaranteed Not guaranteed Not guaranteed Guaranteed
Real-time services Not guaranteed Not guaranteed Not guaranteed Guaranteed

RAN
HSDPA 4 User Plane
4-9
4.4 HARQ
The main purpose of introducing HARQ is to reduce the retransmission delay and improve
the retransmission efficiency. HARQ enables fast retransmission at the physical layer. Before
decoding, the UE combines the retransmitted data and the previously received data, thus
making full use of the data transmitted each time. In addition, HARQ can fine-tune the
effective rate to compensate for the errors made by TFRC section.
4.4.1 HARQ Retransmission Principles
The HARQ process of HSDPA involves only the NodeB and the UE, without involving the
RNC. After receiving a MAC-hs PDU sent by the NodeB, the UE performs a CRC check and
reports an ACK or NACK on the HS-DPCCH to the NodeB:
If the UE reports an ACK, the NodeB transmits the next new data.
If the UE reports an NACK, the NodeB retransmits the original data. After receiving the
data, the UE performs soft combining of this data and the data received before, decodes
the combined data, and then reports an ACK or NACK to the NodeB.
RLC retransmission on the DCH involves the RNC, and therefore the RTT is relatively long.
In comparison, HARQ involves only the physical layer and MAC-hs of the NodeB and those
of the UE, and therefore the RTT is reduced to only 6 TTIs.
After a transmission, the HARQ process must wait at least 10 ms before it can transmit the
next new data or retransmit the original data. Therefore, to improve transmission efficiency,
other HARQ processes can transmit data during the waiting time. A maximum of six HARQ
processes can be configured in each of the NodeB HARQ entity and the UE HARQ entity.
Note that not all UE categories support six HARQ processes. For example, the UEs of some
categories can receive data every one or two TTIs. Thus, only two or three HARQ processes
can be configured. The RAN can automatically choose the most appropriate configuration
based on UE capability.
Figure 4-2 HARQ retransmission principle

4 User Plane
RAN
HSDPA
Issue 02 (2009-06-30)
4.4.2 Soft Combining During HARQ
Before decoding a MAC-hs PDU, the UE performs soft combining of all the data received
before to improve the utilization of Uu resources and thus increase the cell capacity. The size
of the UE buffer determines the number of coded bits or the size of transport blocks.
For HARQ retransmission between the NodeB and the UE, two combining strategies are
available. They are Chase Combining (CC) and Incremental Redundancy (IR). In the case of
CC, all retransmitted data is the same as previously transmitted data. In the case of IR, the
retransmitted data may be different from the previously transmitted data. In comparison, IR
has a higher gain than CC but requires more buffer space. CC can be regarded as a special
case of IR. The IR strategy is hard-coded in Huawei RAN.
4.4.3 Preamble and Postamble
If the HS-SCCH is received, the UE checks whether the HS-PDSCH is also correctly received
and then reports an ACK or NACK in the first slot of the HS-DPCCH subframe. If the
HS-SCCH is erroneously received, the UE does not report any information in the first slot of
the HS-DPCCH subframe. This type of transmission is called DTX. In the case of high
interference, the NodeB may demodulate DTX as ACK by mistake when demodulating the
HS-DPCCH. Thus, the lost data blocks cannot be retransmitted through HARQ retransmission,
and the reception can be ensured only through RLC retransmission. To meet the requirement
of the 3GPP specifications for a low DTX misjudgment probability, more power has to be
allocated for HS-DPCCH ACK/NACK.
To solve this problem, 3GPP TS 25.214 introduces preamble and postamble. When the NodeB
demodulates an HS-DPCCH ACK/NACK, it considers the subframe prior to and the subframe
next to the HS-DPCCH subframe in addition to the HS-DPCCH subframe itself. Thus, for a
certain DTX misjudgment probability, the introduction of preamble and postamble reduces
the power required by ACK/NACK, lower the downlink load level, and increase the uplink
capacity. "HS-DPCCH Preamble Support" is an optional feature.
Figure 4-3 HS-DPCCH preamble and postamble

RAN
HSDPA 4 User Plane
4-11
4.5 TFRC Selection
The TFRC selection algorithm handles the MAC-hs queues in descending order of their
priorities determined by the scheduler. The main tasks of the algorithm for each queue in each
TTI are as follows:
Determining the amount of data that can be transmitted by the queue
Determining the modulation scheme of the queue
Allocating appropriate power and channelization codes to the queue
During the handling, the TFRC selection algorithm considers the following factors:
Channel conditions of the UE
Available resources
Amount of data buffered in the MAC-hs queue
Based on these factors, the algorithm allocates appropriate resources and selects appropriate
transport block sizes to ensure the transmission quality and avoid wasting the resources.
When the channel conditions are bad, the algorithm selects small transport block sizes to
ensure that the data is received correctly and transmitted continuously. When the channel
conditions are good, the algorithm selects large transport block sizes for higher transmission
rates and QoS.
4.5.1 Basic Procedure of TFRC Selection
The basic procedure of the TFRC selection algorithm is as follows:
Step 1 Based on the CQI reported by the UE, available power, and available channelization codes,
the algorithm searches a CQI mapping table for the TBSmax, that is, the maximum MAC-hs
transport block size (TBS).
For details, see section 4.5.2 "Determining the TBSmax."
Step 2 Based on the TBSmax and the amount of data buffered in the queue, the algorithm determines
the most appropriate MAC-hs TBS (TBSused). Here, TBSused <= TBSmax. Based on the TBSused,
the algorithm determines the most appropriate power, codes, and modulation scheme.
For details, see section 4.5.3 "Determining the TBS , Modulation Scheme, Power, and
Codes
used
."
Step 3 Based on the TBSused, the algorithm calculates the number of MAC-d PDUs that can be
transmitted in the MAC-hs PDU.
For details, see section 4.5.4 "Determining the Number of MAC-d PDUs."
Step 4 The algorithm updates the records of the remaining power, codes, and HS-SCCH quantity. If
all of these resources are available and another MAC-hs queue is waiting for resource
allocation, then the algorithm repeats the previous steps. Otherwise, the algorithm ends the
handling.
----End
4.5.2 Determining the TBSmax
The UE assumes that the transmit power of the HS-PDSCH on the network side is as follows:

4 User Plane
RAN
HSDPA
Issue 02 (2009-06-30)
Δ+Γ+=− CPICHPDSCHHS PP
where
PCPICH is the transmit power of the CPICH.
is the measurement power offset (MPO). It is specified by the parameter
HsPdschMPOConstEnum on the RNC side and sent to the NodeB and UE.
Γ
is the reference power adjustment. It is set to 0 in most cases. For details, see 3GPP
TS 25.214.
Δ
On this assumption, the UE reports the CQI through the HS-DPCCH to the NodeB. The CQI
indicates the channel conditions of the UE. A higher CQI indicates that the channel quality is
better and therefore the NodeB can send a larger MAC-hs transport block to the UE.
The NodeB creates a CQI mapping table for each UE category. For each CQI, this table
provides a corresponding MAC-hs TBS and a modulation scheme based on the assumed
power ( Δ+Γ+=− CPICHPDSCHHS PP ) and the number of channelization codes. Such
combinations ensure that the block error rate (BLER) of MAC-hs transport blocks on the Uu
interface does not exceed 10%. The table is obtained on the basis of many simulations and test
experiences. It plays a very important role in HSDPA resource allocation.
If the available power of the HS-PDSCH is higher than the assumed power, a larger MAC-hs
TBS is allowed, which is equal to the TBS corresponding to the adjusted CQI. The adjusted
CQI is calculated as: reported CQI + (available power - assumed power). In this way, the
algorithm provides higher transmission rates.
If the available power is lower than the assumed power, the supported MAC-hs TBS needs to
be reduced to the one corresponding to the adjusted CQI. The adjusted CQI is calculated as:
reported CQI – (assumed power - available power). In this way, the algorithm ensures
transmission correctness.
Thus, the algorithm can determine the TBSmax of the UE in the current cell after obtaining the
CQI reported by the UE, available power, and available codes.
Here is an example. Assume that the CQI reported by the UE is 5, the available power is equal
to the assumed power, and the number of available codes is 4. Then, the TBSmax is 3,762 bits
and the modulation scheme is QPSK. The following figure shows this example.

RAN
HSDPA 4 User Plane
4-13
4.5.3 Determining the TBSused, Modulation Scheme, Power, and
Codes
If the data buffered in the MAC-hs queue is much enough to fill the space for carrying data in
a transport block with the TBSmax, then the TBSmax is taken as the TBS to be used (TBSused).
Accordingly, the modulation scheme corresponds to this TBS is taken as the one to be used.
The algorithm then determines the power and channelization codes to be used, according to
the method mentioned in section 4.5.2 "Determining the TBSmax."
The TBSmax, however, may be much larger than the data buffered in the MAC-hs queue. If
this TBS is used, too many padding bits reduce the spectrum efficiency. To solve this problem,
the algorithm searches the CQI mapping table backward for the CQI or the number of codes
so as to obtain the most appropriate TBS and the corresponding modulation scheme. This
TBS should be the smallest one in the TBS set that can carry the buffered data. The power and
code resources determined through backward searching are taken as the ones for allocation.
Huawei supports three backward-searching methods, which are specified by the parameter
RscAllocM on the NodeB side:
If the parameter is set to Code_Pri, the algorithm prefers the use of codes. Under the
precondition that the transport block with the TBS is large enough to carry the buffered
data, the algorithm first reduces the power. If the corresponding CQI decreases to the
smallest one but the precondition is still met, the algorithm attempts to reduce the
number of codes. This setting is applicable the outdoor macro base station with limited
power.
If the parameter is set to Power_Pri, the algorithm prefers the use of power. Under the
precondition that the transport block with the TBS is large enough to carry the buffered
data, the algorithm first reduces the number of codes. If the number of codes decreases to
1 but the precondition is still met, the algorithm attempts to reduce the power. This
setting is applicable to indoor application with limited codes.
If the parameter is set to PowerCode_Bal, the algorithm balances the use of power and
the use of codes. Under the precondition that the transport block with the TBS is large
enough to carry the buffered data, the algorithm reduces the power and codes in a
balanced mode. This setting protects the codes or power from being used up, thus
improving the resource usage and increasing the cell capacity.
The following figure shows the backward-searching methods used when the parameter is set
to Code_Pri or Power_Pri.
The following figure shows the backward-searching methods used when the parameter is set
to PowerCode_Bal.

4 User Plane
RAN
HSDPA
Issue 02 (2009-06-30)
4.5.4 Determining the Number of MAC-d PDUs
TBSused is the used MAC-hs PDU size. It contains a MAC-hs header and the MAC-hs payload.
The size of MAC-hs payload is equal to the total size of MAC-d PDUs. Assume that S
represents (TBSused – (MAC-hs PDU header size))/(MAC-d PDU size). Then, round down S to
the nearest integer to obtain the number of MAC-d PDUs to be transmitted.

RAN
HSDPA 5 QoS and Diff-Serv Management
5-1
5 QoS and Diff-Serv Management
QoS Management
Diff-Serv Management
apping and Configuration
5.1 QoS M
e the
servi The requirements
Q
ires higher service rates to provide better user experience.
es shorter delay to provide
better user experience.
ctions. The following table
l ons bet nd
T ions b n HSDPA functions an QoS indicators
QoS Parameter M
anagement
Th goal of service-oriented QoS management is to improve user experience by reducing
ce delay and BLER and by increasing the service rate and continuity.
for oS vary according to the type of service:
The conversational service (including the CS voice and VoIP) has a relatively high
requirement for service delay and a certain requirement for BLER.
The streaming service has a requirement for guaranteed bit rate (GBR).
The FTP service has a very high requirement for BLER and error-free transmission. In
addition, this service requ
The HTTP service has a high requirement for error-free transmission and a certain
requirement for response delay. In addition, this service requir
HSDPA QoS management is implemented by related HSDPA fun
ists the relati ween HSDPA functions a QoS indicators.
able 5-1 Relat etwee d
Function Service Connectivity Service Delay Service Rate BLER
Mobility
management
HSDPA bearer
ngmappi
Load control

RAN
HSDPA
Issue 02 (2009-06-30)
Function Service Connectivity Service Delay Service Rate BLER
RLC
retransmission
Flow control
Congestio
control
n
HARQ
MAC-hs
scheduling
TFRC selection
Thes
ment
" and the Handover Parameter
ping."
ntrol the access so as to
by the admitted users vary with the changed channel
ad
n and HARQ
ncy, HSDPA
r. HARQ, however, cannot completely ensure
o much data from waiting in the buffer at the
protects service
g from the buffer at the MAC-hs.
ing
e relations between HSDPA functions and QoS indicators are described as follows:
Mobility manage
Service continuity is implemented by mobility management.
For details, see section 3.3 "Mobility Management
Description.
Bearer mapping
HSDPA bearers increase the service rate greatly and reduce the service delay.
For details, see section 3.1 "Bearer Map
Load control
The network resources are limited. Therefore, when a large number of users attempt to
access the network, the access control function is required to co
ensure the QoS of the admitted users.
The network resources consumed
qualities, which may lead to network congestion. To relieve congestion, the overlo
control function is required to ensure the QoS of most users.
For details on load control, see the Load Control Parameter Description.
RLC retransmissio
To achieve error-free transmission and improve transmission efficie
introduces HARQ at the physical laye
error-free transmission. Therefore, it should work with RLC retransmission and TCP
retransmission.
For details, see sections 4.2 "RLC and MAC-d" and 4.4 "HARQ."
Flow control and congestion control
By allocating appropriate Iub bandwidth to users, the flow control function reduces the
transmission time. Thus, it prevents to
MAC-hs and avoids unnecessary RLC retransmissions. In addition, it
data from overflowin
Through congestion detection and congestion control, the congestion control function
reduces the packet loss probability.
For details, see section 4.1 "Flow Control and Congestion Control."
MAC-hs schedul

RAN
5-3
Based on the waiting time, achieved service rate, and GBR, the MAC-hs scheduling
function sorts the users to meet the requirements for transmission delay and transmission
rate on the Uu interface. For details, see section 4.3 "MAC-hs Scheduling."
TFRC selection
, available codes, actual channel quality, and actual data
function selects appropriate transport blocks and modulation
es. For details, see section 4.5 "TFRC Selection."
5.2 Diff-Serv Management
ffe fferent users have different priorities.
During r provided. Differentiated services for
Differentiated services based on user priorities
Diff-Serv management, differentiated services based on SPI
5.2.1 SPI W
user priority and service type. Based on
these .
The
prov
is redundant after the requirements of all
on Resource Management Parameter Description.
d
istent with the proportion of
, see section 4.3.2 "Calculating Priorities."
In this way, a user with a larger SPI weight is provided with better services, and the QoS
figure SPI weights,
Based on the available power
amount, the TFRC selection
schemes to increase data rat
Di rent services have different service types, and di
esource allocation, differentiated services are
HSDPA users are as follows:
Differentiated services based on service types
To further quantify the effect of
weights are introduced. This section describes the differentiated services based on SPI
weights and the differentiated service policies.
eight Description
HSDPA differentiated services consider two factors:
two factors, each type of service is mapped to an SPI, and each SPI is allocated a weight
SPI weight is a percentage ranging from 0% to 100%. Different SPI weights are set to
ide the following differentiated services for different users:
Differentiated services on the Iub interface
If the downlink bandwidth on the Iub interface
BE services for the GBR are met, the redundant bandwidth is allocated to users so that
the proportion of rates among users is consistent with the proportion of SPI weights. For
details, see the Transmissi
Differentiated services on the Uu interface
When the load resources in the uplink on the Uu interface are insufficient to meet the
requirements of HSDPA users for the GBR, a user with a greater SPI weight can obtain a
rate closer to the GBR.
When the load resources in the uplink on the Uu interface are redundant after the
requirements of HSDPA users for the GBR are met, the redundant resources are allocate
to users so that the proportion of rates among users is cons
SPI weights. For details
values of users are measured by proportion. For details about how to con
see section 5.3 "QoS Parameter Mapping and Configuration."
The SPI weight is optional.

RAN
HSDPA
Issue 02 (2009-06-30)
5.2.2 Differ
Differentiated S
ied into delay-sensitive services and
ith service types. For example, for
delay-sensitive services, the EPF scheduling function ensures that the transmission delay of
permissible range. For throughput-sensitive
services, the EPF scheduling function tries to provide rates not lower than the GBR.
cording to service types, the differentiation can be
implemented between throughput-sensitive services.
ities."
Differentiated
at is, a large
bandwidth is always allocated on the Iub interface. For low-priority non-real-time services,
occurs on the Iub interface, the bandwidth
5.2.3 Differ
GBR and HBR
r
heduling and flow control, the basic GBR services
, a higher rate is not required for the user,
thus saving system resources. This rate is defined as HBR. The HBR is configured on the
ers. For details on the GBR and HBR used in HSDPA scheduling,
see sec
On t ted
servi BR used in HSDPA flow control, see the Transmission
so
entiated Services Based on Service Types
ervices on the Uu Interface
In the EPF scheduling function, all services are classif
throughput-sensitive services. The QoS targets vary w
data packets does not exceed the maximum
Delay-sensitive service users have higher requirements for the QoS. Therefore,
delay-sensitive services have higher priorities than throughput-sensitive services.
If the SPI weight is specified ac
For details, see section 4.3.2 "Calculating Prior
Services on the Iub Interface
For high-priority real-time services, Iub flow control is not performed, th
flow control must be performed. When congestion
for these services is decreased.
If the SPI weight is configured according to service types, the differentiation can be
implemented between non-real-time services.
For details, see the Transmission Resource Management Parameter Description.
entiated Services Based on User Priorities
Services Based On User Priorities
When setting up a service, the CN provides an ARP parameter, which specifies the user
priority. Based on user priorities, the CN provides differentiated GBR and HBR services fo
BE services.
When the CN does not configure the GBR for BE services carried on the HSDPA channel,
they cannot obtain resources if the system is congested. To provide basic services for BE
services, the GBR is configured on the RAN side. The GBR based on user priorities is
configured on the RNC. During HSDPA sc
are provided for BE services as much as possible.
Based on actual user experience, when the rate of the BE service reaches a specified value,
the requirements of the user are met. In this case
RNC according to user priorities.
On the Uu interface, HSDPA scheduling considers the GBR and HBR and thus provides
differentiated services for us
tion 4.3.2 "Calculating Priorities."
he Iub interface, HSDPA flow control considers the GBR and thus provides differentia
ces for users. For details on the G
Re urce Management Parameter Description.

RAN
5-5
Differentiated Ser
improving
A services.
When there are redundant system resources after the GBR requirements of all online HSDPA
sources. On the Uu
s are close to the
I
t, see section 4.3.2 "Calculating Priorities" and the
Transmission Resource Management Parameter Description.
5.3 QoS Parameter Mapping and Configuration
D tiated t c R, H
weight. Among these factors, the MB d A
RNC through the RAB assignment me BR,
service can be set on the RNC side.
Table 5-2 QoS parameters description
vices Based on SPI Weights
When the network is congested, differentiated services are provided for users during
bandwidth allocation.
When the basic QoS requirements of some users cannot be met because of system
congestion, high-priority users or services are allocated resources preferentially, thus
improving the user satisfaction.
When the system resources are redundant even after the basic QoS requirements of all
the users are met, the redundant resources are allocated to users additionally.
High-priority users or services are allocated additional resources, thus further
the user satisfaction.
If the SPI weights are configured according to service types, the quantified differentiated
services can be implemented between different types of HSDP
users are met, SPI weights affect the allocation of the redundant system re
interface, the scheduling probabilities among throughput-sensitive service
proportion of their SPI weights. On the Iub interface, the proportion of available Iub
bandwidth among BE services is also close to the proportion of their SPI weights.
Flow control and MAC-hs scheduling are involved in Diff-Serv management based on SP
weights. For details on the SPI weigh
ifferen services managemen onsiders MBR, GB BR, TC, THP, ARP, and SPI
RP are sent by the CN to the
and SPI weight of the BE
R, GBR, TC, THP, an
ssage. The GBR, H
Diff-Serv Full Name Parameter ID Description
Factor
MBR bit rate
SingalDlMBR
StreamUlMBR
StreamDlMBR
BR
ConverDlMBR
maximum SingalUlMBR
ConverUlM
MBR specifies the UL/DL
maximum bit rate of signals
and BE for PS domain users.
ARP Allocation/Retention
priority
ARP1Priority
~ARP14Priority tention priority
4.
0 is invalid, and 15 is known
as no priority,Huaweideal the
same with APR14.
User_priority of
Allocation/Re
1~1

RAN
HSDPA
Issue 02 (2009-06-30)
Diff-Serv Full Name Parameter ID Description
Factor
TC traffic class TrafficClass
e
of
This parameter specifies the
traffic class that the servic
belongs to. BE services are
two classes: interactive and
background.
THP
is
only
traffic handling
priority
THP This parameter specifies the
Traffic Handling Priority
(THP) class that the THP
priority is mapped to. Th
parameter is valid for
interactive services.
SPI scheduling priority
indicator
SPI This parameter indicates the
scheduling priority. The value
15 indicates the highest
priority and the value 0
indicates the lowest.
GBR guaranteed bit rate UlGBR
DlGBR
determined by the NAS.
he
d.
GBR is configured on the
basis of the MAC-hs queue.
For the streaming service, the
GBR specifies the rate that can
meet the requirement for user
experience. The GBR is
For the BE service, the GBR
specifies the required
minimum rate for the service
of the users. The GBR of a BE
service user is set through t
SET USERGBR comman
SPI
Weight
scheduling priority
indicator weight
FACTOR This parameter specifies the
factor associated with the
scheduling priority indicator.
This factor is used to calculate
the step of rate upsizing.
HBR happy bit rate HappyBR
with
different user priorities. The
Happy bit is the private
information element on the Iub
interface and it is used for the
flow control by the NodeB.
The Happy bit rate is the data
rate at the MAC layer.
Happy bit rate is for the
best-effort (BE) service
Figure 5-1 shows the mapping between these factors.

RAN
5-7
Figure 5-1 Mapping between the factors considered in differentiated services management
The mapping can be set on the RNC side:
User Priority decided by ARP, the mapping of ARP to User Priority is set by the SET
USERPRIORITY command.
Table 5-3 Mapping of ARP to User Priority (Gold, Silver, and Copper correspond to user priorities
1, 2, and 3 respectively.)
ARP 0 1 2 3 4 5 6 7 8 9 1
0
1
1
1
2
1
3
14 15
UserPri
ority
Err
or
1 1 1 1 1 1 1 2 2 2 3 3 3 3 3
The mapping of TrafficClass, UserPriority and THP to SPI is set by the SET
SCHEDULEPRIOMAP command in RNC.
Table 5-4 Default mapping of user priority to SPI (Gold, Silver, and Copper correspond to user
priorities 1, 2, and 3 respectively.)
TrafficClass UserPriority THP SPI
1 1 10
1 2 9
1 3 to 15 8
2 1 7
2 2 6
2 3 to 15 5
3 1 4
3 2 3
Interactive
3 3 to 15 2

RAN
HSDPA
Issue 02 (2009-06-30)
TrafficClass UserPriority THP SPI
1 None 8
2 None 5
Background
3 None 2
SPI 0 and SPI 1 are not used.
The mapping of SPI to FACTOR is set on the RNC through the SET SPIFACTOR
command. Though the SPI configuration considers user priorities and service types, the
SPI weight can also be configuration according to user priorities.
Table 5-5 Default setting of algorithm based on SPI
SPI Weight of SPI (Factor)
15 100%
14 100%
13 100%
12 100%
11 100%
10 100%
9 100%
8 100%
7 90%
6 90%
5 90%
4 80%
3 80%
2 80%
The mapping of TrafficClass, THPClass and UserPriority to GBR is set by the SET
USERGBR command in RNC.
The mapping of TrafficClass and UserPriority to HappyBR is set by the SET
USERHAPPYBR command in RNC.

RAN
HSDPA 6 Parameters
6-1
6 Parameters
The following describes the parameters related to HSDPA.
parameter, see Table 6-1. For the default value, value ranges, and
MML commands eter, seeTable 6-2.
T arameter descripti
For the meaning of each
of each param
able 6-1 HSDPA p on(1)
Parameter ID Description
AllocCodeMode If Manual is chosen, allocating [Code Number for HS-PDSCH] the equal of
cating
um code
configured HS-PDSCH code number. If Automatic is chosen, allo
HS-PDSCH code number between configured HS-PDSCH Maxim
number and HS-PDSCH Minimum code number. At the earl
ARP10Priority User_priority corresponding to Allocation/Retention priority 10.

6 Parameters
RAN
HSDPA
Issue 02 (2009-06-30)
CellLdrSfResThd Cell SF reserved threshold. The code load reshuffling could be triggered
only when the minimum available SF of a cell is higher than this thresho
The lower the code resource LDR trigger threshold is, the easier the
downlink code resource enters the ini
ld.
CodeAdjForHsdpaSwitch Code reshuffle switch based on H. If the switch is enabled, code occupie
by the R99 service is adjusted toward codes with small numbers. When
[Allocate Code Mode] is set to Automatic, code can be used by HSDPA
increases and HSDPA throughput is
d
improved.
CodeAdjForHsdpaUserNumThd
number of users on the tree to be reshuffled is no greater
H-based code tree reshuffle user number threshold. When the switch is
enabled, if the
than this parameter, the reshuffle is allowed. Otherwise, the reshuffle is
given up. This parameter limits the
ConverDlMBR versation for PSThis parameter specifies the DL maximum bit rate of con
domain user.
ConverUlMBR This parameter specifies the UL maximum bit rate of conversation for PS
domain user.
DlGBR This parameter specifies the DL GBR of the BE service.
DynCodeSw Dynamic Code Resource Distribuiton Switch of HSDPA
FACTOR rity
alculate the step of rate upsizing.
This parameter specifies the factor associated with the scheduling prio
indicator. This factor is used to c
HappyBR This parameter specifies the Happy bit rate of the best-effort (BE) service
with different user priorities. The Happy bit rate is the private information
element on the Iub interface and it is used for the flow control by the
NodeB. When resource is limit
HspaPower This parameter specifies the difference between the total HSPA power and
the maximum transmission power of a cell. The maximum value of HSPA
dynamical power can be adjusted to the total amount of HSPA power.
parameter value is set too low, the tota
If the
HsPdschCodeNum The parameter specifies the number of HS-DPSCH codes. This parameter is
valid only when "Allocate Code Mode" is set to "Manual". If the parameter
value is set too low, the HSDPA code resources are restricted and the
HSDPA performance is affect. If the par
HsPdschMaxCodeNum The parameter determines the maximum number of HS-PDSCH codes
(SF=16). This parameter is valid only when "Allocate Code Mode" is set to
"Automatic". The number of codes used by the HS-PDSCH is dynamically
set between "Code Max Number for HS-PDSCH" and "Co
HsPdschMinCodeNum The parameter specifies the minimum number of the HS-PDSCH codes
(SF=16). This parameter is valid only when "Allocate
Automatic. The number of codes used by the HS-PDSCH is dynamically
set between "Code Max Number for HS-PDSCH" and "C
Code Mode" is set to

RAN
HSDPA 6 Parameters
6-3
HsPdschMPOConstEnum
nstant)).
CQI in some scenarios w
Measure Power Offset (MPO) Constant is used to compute Measure Power
Offset, as shown in Measure Power Offset = Max(-6,
Min(13,CellMaxPower - PcpichPower - Measure Power OffsetCo
If the parameter value is unreasonable, the
HsScchCodeNum This parameter decides the maximum number of subscribers that the Node
can schedul
B
e in a TTI period. In the scenarios like outdoor macro cells with
power restricted, it is less likely to schedule multiple subscribers
simultaneously, so two HS-SCCHs are c
MaxNonConverHarqRt on-Conversational serive in CellMax HARQ Retransmission Times of N
DCH state
PwrMgn Power Margin Ratio, to prevent the total power from exceeding the 100%
power margin in 2 ms.
RscAllocM Resource Allocate Method of HSDPA
RscLmSw Resource Limiting Switch of HSDPA
SingalDlMBR maximum bit rate of signal for PS domainThis parameter specifies the DL
user.
SingalUlMBR This parameter specifies the UL maximum bit rate of signal for PS domain
user.
SM Scheduling Method of HSDPA
SPI This parameter indicates the scheduling priority. The value 15 indicates
highest priority and the value 0 indicates the lowest.
the
StreamDlMBR This parameter specifies the DL maximum bit rate of streaming for PS
domain user.
StreamUlMBR This parameter specifies the UL maximum bit rate of streaming for PS
domain user.
THP
ed on the logical channel. The value 1 means the highest priority,
This parameter specifies the Traffic Handling Priority (THP) of each traffic
class carri
the value 14 means the lowest priority, and the value 15 means no priority.
THPClass the
s mapped to. This parameter is valid for only interactive
This parameter specifies the Traffic Handling Priority (THP) class that
THP priority i
services.
TrafficClass This parameter specifies the traffic class that the service belongs to. Based
on Quality of Service (QoS), there are two traffic classes: interactive,
background.
UlGBR This parameter specifies the UL GBR of the BE service.
USERPRIORITY This parameter specifies the user priority. The user classes in descending
order of priority are Gold, Silver, and then Copper.
8KRSCLMT Upper limit ratio of the power for the user with 8 kbps GBR to the total
power of the cell

60118029 hsdpa-parameter-description-130219192019-phpapp01

60118029 hsdpa-parameter-description-130219192019-phpapp01

Recommended

Recommended

More Related Content

Viewers also liked

Viewers also liked (6)

Similar to 60118029 hsdpa-parameter-description-130219192019-phpapp01

Similar to 60118029 hsdpa-parameter-description-130219192019-phpapp01 (20)

More from Hatim100

More from Hatim100 (20)

Recently uploaded

Recently uploaded (20)

60118029 hsdpa-parameter-description-130219192019-phpapp01