ZTE UMTS UR14 NodeB HSDPA Packet Scheduling Feature Guide

NodeB HSDPA Packet
Scheduling
Feature Guide
NodeB HSDPA Packet Scheduling

Version
V1.00
Date
2015/03/30
Author
DuJuan,
Xuyun
Reviewer
Wangrui
Notes
Compared with UR13, chapter 6 was

modified.
2015 ZTE Corporation. All rights reserved.

ZTE CONFIDENTIAL: This document contains proprietary information of ZTE and is not to be disclosed or used
without the prior written permission of ZTE.
Due to update and improvement of ZTE products and technologies, information in this document is subjected to
change without notice.
ZTE Confidential Proprietary
TABLE OF CONTENTS
1
Feature Attribute ............................................................................................... 5
2
2.1
2.1.1
2.1.2
2.2
2.3
Overview ............................................................................................................ 5
Feature Introduction ............................................................................................. 6
ZWF23-01-013 HSDPA Fast Scheduling ............................................................. 6
ZWF23-05-002 HSDPA Flow Control ................................................................... 6
License Control .................................................................................................... 7
Correlation With Other Features .......................................................................... 7
3
3.1
3.1.1
3.1.2
3.1.3
3.2
3.2.1
3.2.2
3.3
3.3.1
3.3.2
3.3.3
3.3.4
3.3.5
3.3.6
3.3.7
3.3.8
3.3.9
3.3.10
3.3.11
Technical Description ....................................................................................... 8

ZWF23-01-013 HSDPA Fast Scheduling ............................................................. 8
HSDPA Resource Allocation Policies ................................................................... 8
HSDPA Scheduling Algorithms ............................................................................ 9
HSDPA TFRC Selection Algorithm .................................................................... 14
ZWF23-05-002 HSDPA Flow Control ................................................................. 19
HSDPA Flow Control Algorithm.......................................................................... 19
Measurement of HS-DSCH Required Power...................................................... 25
Impact of HSPA+ on the Scheduler.................................................................... 27
Impact of 64QAM Modulation on the Scheduler ................................................. 27
Impact of MIMO on the Scheduler ...................................................................... 30
Impact of DC-HSDPA on the Scheduler ............................................................. 33
Impact of Enhanced CELL-FACH on the Scheduler ........................................... 33
Impact of CPC on the Scheduler ........................................................................ 34
Impact of CS Voice Service over HSPA on the Scheduler ................................. 35
Impact of DC+MIMO+64AM on the Scheduler ................................................... 35
Impact of Dual-Band DC-HSDPA on the Scheduler ........................................... 35
Impact of 4C-HSDPA on the Scheduler ............................................................. 35
Impact of DB DC-HSDPA+MIMO on the Scheduler ........................................... 36
Impact of 4C-HSDPA+MIMO on the Scheduler .................................................. 36
4
4.1
4.2
Parameters....................................................................................................... 36
ZWF23-01-013 HSDPA Fast Scheduling ........................................................... 36
5
5.1
5.1.1
5.1.2
5.2
5.2.1
Related Counters and Alarms ........................................................................ 38

Related Counters ............................................................................................... 38
Related Alarms .................................................................................................. 47
5.2.2
6
6.1
6.2
6.2.1
6.2.2
6.3
6.3.1
6.3.2
6.4
6.4.1
6.4.2
6.5
6.5.1
6.5.2
Engineering Guide .......................................................................................... 47

Application Scenario .......................................................................................... 47
Feature Activation Procedure ............................................................................. 47
Feature Validation Procedure............................................................................. 50
Feature Deactivation Procedure......................................................................... 51
Impact on the Network ....................................................................................... 52
Abbreviation .................................................................................................... 52
Reference Document....................................................................................... 54
FIGURES
Figure 3-1 Resource Allocation During UE Scheduling ......................................................17
Figure 3-2 Flow Control Procedure ....................................................................................21
Figure 3-3 Leaky Bucket Flow Control Principle .................................................................22
Figure 6-1 Configure HSPA Support Method of the cell .....................................................48
Figure 6-2 Configure HSDPA Function Status .................................................................48
Figure 6-3 Configure HSDPA Function Status .................................................................49
Figure 6-4 Configure the number of HS-PDSCH ................................................................50
TABLES
Table 2-1 License Control List ............................................................................................ 7
Table 3-1 WeightofDelay Calculation .................................................................................12
Table 3-2 HS-PDSCH Starting Code Numbers ..................................................................28
Table 4-1 HSDPA Fast Scheduling OMMB Parameters List..............................................36
Table 4-2 HSDPA Fast Scheduling OMMR Parameters List .............................................37
Table 5-1 HSDPA Fast Scheduling Counter List ................................................................38
Feature Attribute
RNC version: [RNC V3.14.10, V4.14.10]
NodeB version: [ZXSDR V4.14.10.10]
Attribute: [Optional]
Involved NEs:
NE Name
Related or Not
Special Requirement
MS/UE
None
BTS/Node B
None
BSC/RNC
None
iTC
None
MSC
None
MGW
None
SGSN
None
GGSN
None
HLR
None
: involved
-: not involved
Overview
After the High Speed Downlink Packet Access (HSDPA) technology is introduced to the
WCDMA system, the MAC-hs layer is added to both the Node B and UE. The NodeB
HSDPA provides the following functions:
2msTTI.
Performs cell-based fast scheduling for multiple UEs.
The Adaptive Modulation and Coding (AMC) provides the HS-PDSCH air interface
transmission format (number of channelization codes, modulation mode, and TB
size).
Transmits HS-SCCH control information and HS-DSCH data information.
Demodulates messages such as ACK, NACK and CQI carried on the HS-DPCCH.
Performs downlink HS-DSCH flow control.
2.1
Feature Introduction
2.1.1
ZWF23-01-013 HSDPA Fast Scheduling

Fast scheduling procedures are described as below:
Receive FP data via Iub port, save it in the buffer of each UE.
First, select a UE, which has data to be sent, according to the scheduling priority
algorithm. Then, the HARQ entity selects an idle HARQ process to transmit the data.
The scheduling priority algorithm mainly includes Round Robin (RR), Maximum
Carrier to Interface (Max C/I) and Proportional Fairness (PF).
Select a suitable HS-PDSCH air-interface transmission format according to the UE

channel quality, UE category, available power, code resources, and other
information. Then send the data via the air interface.
2.1.2
ZWF23-05-002 HSDPA Flow Control

The air interface rate of an HSDPA UE depends on several factors including UE channel
quality, number of UEs to be scheduled in a cell and code/power resources of a cell.
These factors are changeable. The scheduler uses flow control to administrate the
incoming rate of downlink HS-DSCH data frames of each UE so as to make it consistent
with the outgoing rate at the air interface. This can avoid a situation that UEs have too
much or too little data amount to be sent at Node B.
At the Iub interface, the HSDPA scheduler informs the RNC to control the transmission
rate of a UE by sending Capacity Allocation Frame. If a UE fails to receive CAF for a
specified time, the RNC will proactively send Capacity Request Frame to the scheduler.
The HSDPA scheduler will send a CAF to the RNC in response.
The UE downlink rate, which is sent from the RNC to Node B scheduler, will not exceed
the rate specified in CAF.
2.2
License Control
This feature is under license control. Before this feature is commissioned, it must have
authorization on the OMC or LMT.
Table 2-1
License Control List
Feature ID
Feature Name
License
Configured
Control Item
NE
Unit
WCDMA
ZWF23-01-A
HSDPA
Package
Introduction
HSDPA
Introduction
RNC
CELL
Package
2.3
Correlation With Other Features

1. Required Features
None
2. Mutually Exclusive Features
None
3. Affected Features
None
Technical Description
3.1
3.1.1
HSDPA Resource Allocation Policies

The HSDPA resources fall into two types: code and power. The following sections
describe the two HSDPA resource allocation policies:
3.1.1.1
HSDPA Power Resource Allocation Scheme

HSDPA power allocation refers to the algorithm how the HSDPA maximum available
power is allocated.
The HSDPA power resource allocation policies can be configured with the parameter
HSPA Total Downlink Power Allocation Method (HsdschTotPwrMeth). There are two
types of HSDPA power resource allocation policies: RNC dynamic allocation algorithm
and Node B allocation algorithm.
For detailed information about the RNC dynamic allocation algorithm, refer to the ZTE
UMTS Power Control FG.
The Node B allocation algorithm is implemented by the Node B. The RNC only needs to
configure the HSDPA available power to be 100%.
Because the downlink power control of R99 is implemented in the RNC, the Node B
cannot control the power of R99 UEs The Node B allocation algorithm supports the
function of proactively bypassing the R99 channel downlink transmit power, that is, the
maximum available power of related HSPA downlink channels must not exceed a certain
value. The value is obtained by subtracting the common channel power and downlink
transmit power of R99 channels from the maximum cell transmit power. The proactive
bypass function enables cell power sharing among HSPA and R99 UEs.
The procedure that the HSPA scheduler proactively bypasses the R99 channel downlink
transmit power is described as follows:
In each 2 ms TTI, the cell TSSI reported through RF minus the power consumed by
HSDPA and HSUPA is the R99 power and common channel power at the moment. The
available HSPA power in the next 2 ms TTI is obtained by subtracting the R99 power and
common channel power from the maximum cell transmit power.
In each 2 ms TTI, the current HSDPA maximum available power is obtained by
subtracting the HSUPA power used during this TTI from the available HSPA power
calculated in the previous TTI.
3.1.1.2
HSDPA Code Resource Allocation Scheme

HSDPA code resource allocation refers to the algorithm how HS-PDSCH maximum
available channelization codes are allocated.
HS-PDSCH uses SF=16 to carry downlink HSDPA data. Each cell can be configured
with a maximum of 15 HS-PDSCH codes. Two policies are used to allocate HSDPA code
resource: RNC algorithm and Node B algorithm. The two policies are controlled by the
parameter HS-PDSCH Code NodeB Assignment Support Indicator (HSNBASSIND).
For detailed information about the RNC allocation algorithm, refer to the ZTE UMTS
Code Resource FG.
The Node B allocation algorithm is described as follows:
The Node B scheduler counts how many SF=16 codes are occupied during each TTI.
Then, the maximum number of available HS-PDSCH codes is obtained by subtracting
the number of occupied SF=16 codes from 15 (the total number of HS-PDSCH codes per
cell).
3.1.2
HSDPA Scheduling Algorithms

The purpose of scheduling algorithms is to calculate the relative priority of all UEs in each
2 ms TTI, prioritize them, and schedule those with higher priorities first. The commonly
used scheduling algorithms include MAX-C/I, RR and PF algorithms.
The MAX C/I algorithm focuses on the maximum throughput of a cell. The RR algorithm
gives equal scheduling opportunity for all UEs. The PF algorithm is a tradeoff of MAX C/I
and RR algorithms, and it is most widely used among vendors and operators.
Among these three algorithms, only the PF algorithm takes into account the HSDPA
service attributes.
For class-I/-B and SRB over HS-DSCH services, the QoS is guaranteed by setting
different Schedule Priority Indicators (SPIs) in the RNC.
For Guaranteed Bit Rate (GBR) stream services, the PF algorithm preferentially
schedules UEs with GBRs not satisfied.
Note: UEs with data to be retransmitted must be scheduled preferentially no matter which
scheduling algorithm is used.
The following sections describe the principles of the three scheduling algorithms:
3.1.2.1
MAX C/I Algorithm

The MAX C/I algorithm only takes into account channel quality to maximize cell
throughput. The relative priority is calculated in accordance with the following formula:
RelativePriority = CQI * TBSIZE
The Channel Quality Indicator (CQI) is fed back by the HS-DPCCH of the UE. The
maximum MAC-hs Transmission Block Size (TBSIZE) of the UE is obtained by querying
the CQI mapping table for UE categories provided by TS 25.214 based on the current
CQI, UE category, and the number of available HS-PDSCH channelization codes.
3.1.2.2
RR Algorithm
The relative priority of RR algorithm is calculated in accordance with the following
formula:
RelativePriority = Current Time Last Time of UE Scheduling
10
The time unit in the above formula is TTI 2 ms. Current Time refers to the current
scheduling time. Last Time of UE Scheduling refers to the last time when the UE is
scheduled.
It is obvious that the RR algorithm has the longest scheduling waiting time.
3.1.2.3
PF Algorithm
The PF algorithm takes into account both channel quality and historical traffic. In other
words, the PF algorithm takes into account both cell throughput and user fairness. As a
tradeoff between fairness and cell throughput, the PF algorithm is used by default in most
cases. The relative priority of PF algorithm is calculated in accordance with the following
formula:
RelativePriority =WeightofSPI Rate WeightofDelay (1+HistoryFlux)

The Schedule Priority Indicator (SPI) refers to the UE scheduling priority configured
through NBAP signaling, and ranges from 0 to 15. It is related to the services used by the
UE.
WeightofSPI refers to the weight obtained through SPI mapping configured through
the parameter HSDPASPI. The values of SPI_0, SPI_1SPI15 range from 1 to 2000.
The value of SPI_n must be less than the value of SPI_n+1 (n = 0, 1, 214) because
the mapping relationship between SPI and WeightofSPI must be consistent, otherwise,
the SPIs configured at the RNC will be different from the actual scheduling priorities. The
recommended default values of SPI_0, SPI_1SPI15 are set to [10, 12, 14, 17, 21, 25,
30, 36, 43, 52, 62, 74, 89, 107, 128, 154] by default. It is not suggested to randomly
configure SPI_0 SPI_1SPI15 in a way not compliant to the request of SPI_n being
less than SPI_n+1, because it may cause that the SPIs configured at the RNC are
different from the actual scheduling priorities, which then results in network service
priority chaos.
Generally, Rate represents the current instant rate. The calculation of Rate
is different
from that of traditional UEs, MIMO UEs, and MC UEs (Multi-Carrier UE). Rate is
calculated in accordance with the following formula:
For traditional UEs and MIMO single stream UEs :
11
Rate=w1(CQI_s) TBSIZE(CQI_s)
For MIMO dual stream UEs:
Rate=w2(CQI_1) TBSIZE(CQI_1)+ w2(CQI_2) TBSIZE(CQI_2)

For MC-HSDPA UEs: (M=2, DC; M=4, 4C)
Rate=w1(CQI_n) TBSIZE(CQI_n)
CQI_s is the CQI reported by a traditional UE or an MIMO UE under single stream
scheduling. CQI_1 and CQI_2 is the CQI for the primary stream and the secondary
stream reported by an MIMO UE under double stream scheduling. CQI_n is the CQI for
related carrier reported by an MC-HSDPA UE. TB(CQI_s), TB(CQI_1), TB(CQI_2)
and TB(CQI_n) are obtained by querying the CQI mapping table for UE categories
provided by TS 25.214 based on the current CQI.
w1(CQI_s) and w2(CQI_1) refer to the weight obtained through CQI mapping. w1 is
for single stream mapping, w2 is for dual stream mapping.
WeightofDelay refers to the delay weight factor for VoIP and CS Voice Service over
HSPA. The following table describes how to calculate WeightofDelay, where
represents the waiting time of the first packet data of Que in the NodeB,
HOL
Tdiscard
represents the service discardtime configured by the RNC.
Table 3-1 WeightofDelay Calculation
Tdiscard HOL ( ms)
WeightofDelay
< 10
65
[10, 20)
50
[20, 30)
37
[30, 40)
26
[40, 50)
17
[50, 60)
10
12
[60, 70)
[70, 80)
80
The historical flux of a UE is calculated every 2 ms in accordance with the following

formula:
For traditional UEs:
HistoryFlux(n) = HistoryFlux (n-1) * 0.96 + TBSIZE, where, TBSIZE is a
variable because the data volume scheduled each time varies. n refers to the number of
historical scheduling. HistoryFlux(n) refers to the historical flux after n times of
scheduling. TBSIZE refers to the TBSIZE of the last scheduling.
For MIMO UEs:
HistoryFlux(n) = HistoryFlux (n-1) * 0.96 + TBSIZE1 + TBSIZ2, where,
TBSIZE1 and TBSIZ2 refer to the transmit block size of the primary stream and the
secondary stream of MIMO. If the primary stream or the secondary stream is not
scheduled or retransmitted in this TTI, TBSIZE1 or TBSIZ2 is 0.
For MC UEs:
HistoryFlux(n) = HistoryFlux (n-1) * 0.96 + TBSIZE1 + TBSIZ2 + +
TBSIZEm, where, TBSIZE1 , TBSIZ2 and TBSIZEm refer to the transmit block size of the
primary stream and the secondary stream of MIMO. If it is not scheduled or retransmitted
in this TTI in the related carrier m, TBSIZEm is 0.
If a UE is a stream class user, the Guaranteed Bit Rate(GBR) is configured by the RNC.
If a UE is an I/B class user, a minimal GBR, named Nominal Bit Rate (NBR), is
configured by the RNC,
and generally the NBR value is 16K. Both GBRs are sent to the
scheduler through the MAC-hs Guaranteed Bit Rate IE of NBAP. To distinguish stream
class services and I/B class services, it is defined that if the Discard Timer IE of NBAP is
configured, the scheduler considers the UE as a stream class user; if the Discard Timer
IE is not configured, the scheduler considers the UE as an I/B class user.
13
The HSDPA scheduler first schedules stream class UEs, then schedules I/B class UEs
whose GBRs do not satisfy the requirement, and lastly schedules UEs whose GBRs
satisfy the requirement and non-GBR UEs.
For UEs whose GBRs do not satisfy the requirement, they are prioritized and scheduled
in accordance with the degree of GBR not being satisfied.
The non-GBR UEs and UEs whose GBRs satisfy the requirement are prioritized and
scheduled in accordance with their relative priorities obtained by using the PF algorithm.
3.1.3
HSDPA TFRC Selection Algorithm

When selecting several appropriate UEs for scheduling in each TTI, the scheduler needs
to determine the Transport Format and Resource Combination (TFRC) required for each
UE, for example, TBSIZE, HS-PDSCH code information (starting code number and the
number of channelization codes), HS-PDSCH power, HS-SCCH code number, and
HS-SCCH power.
For UEs with data to be retransmitted, the TFRC is selected based on the principle that
TBSIZE, number of HS-PDSCH channelization codes and power, and HS-SCCH power
are kept unchanged.
For UEs with new data to be transmitted, the TFRC is selected starting from UEs with
higher relative priorities until either the code resources or power resources are used up in
the cell.
3.1.3.1
HS-SCCH Code and Power Selection

The selection of HS-SCCH channelization codes is comparatively simple. According to
the protocol, if a UE performs two TTI scheduling consecutively, it must use the same
HS-SCCH code. In other cases, the system can randomly select an HS-SCCH for the
UE.
There are two types of HS-SCCH power control algorithms. One is the constant power
algorithm (namely, 1/4 pilot power is used constantly), and the other is the CQI-based
outer loop power control algorithm. The constant power algorithm is used for test
14
purpose. By default, the CQI-based outer loop power control algorithm is used, and its
principle is described as follows:
The power selection of HS-SCCH is calculated in accordance with the following formula:
PHSSCCH PCPICH Es / Nohsscch 9 Es / Nohspdsch Pdelta

Where,
PCPICH refers to the receive power of the pilot channel (unit: dBm).
Es/Nohsscch is constantly 1.2 dB.
refers to the Measurement Power Offset (MeasPwrOffset) configured for NBAP
signaling.
Es/Nohspdsch = -4.5 + CQI dB.
Pdelta refers to the value obtained based on the HS-SCCH BLER outer loop
adjustment.
The maximum HS-SCCH power cannot be more than the pilot power, and the minimum
HS-SCCH power cannot be less than -16dB of the pilot power.
The measurement power offset (MeasPwrOffset) is configured according to the actual
use of 64QAM, MIMO and DC-HSDPA in the cell. The value of MeasPwrOffset is
indexed by app64QamInd, appMimoInd, and appDcHsdpaInd.
3.1.3.2
TBSIZE, Number of HS-PDSCH Channelization Codes, Modulation, and

Power Selection
Two types of HS-DSCH power control algorithms are provided. One is the average
power control algorithm (namely, the available power is divided equally by all UEs
scheduled in one TTI), and the other is the MPO power control algorithm. The average
power control algorithm is used for test purposes. By default, the MPO power control
algorithm is used.
The MPO power control algorithm is described as follows:
15
The TBSIZE, number of HS-PDSCH channelization codes, modulation mode, and power
selection are closely related to CQI, and the target BLER of a UE at the time when CQI is
generated is 10%. However, the CQI reported by various manufacturers are inaccurate
due to implementation differences or measurement errors, and must be corrected. The
UE reports CQI and ACK information, then the Node B decodes the two values and
adjusts the target CQI in accordance with the ACK value, in other words, the Node B
performs outer loop adjustment for the CQI reported by the UE to minimize the impact
due to measurement errors and
implementation differences among different
manufacturers.
The HSDPA CQI adjustment algorithm is described as follows:
The CQI offset of a UE is initialized to 0. The MAC-hs TB decoding results are
accumulated. The CQI offset of UE increases by 0.01 every time the ACK signal is
detected, and decreases by 0.09 every time the NACK signal is detected. The adjusted
CQI is obtained by adding the CQI reported by the UE to the CQI offset of the UE.
The TBSIZE, number of channelization codes, modulation mode, and reference power
adjustment can be obtained by querying the CQI mapping table for UE categories
provided in TS 25.214 based on the adjusted CQI.
PHSPDSCH PCPICH
Where,
PCPICH refers to the receive power of the pilot channel (unit: dBm).
refers to the Measurement Power Offset (MeasPwrOffset) configured for NBAP
signaling.
refers to the reference power adjustment obtained after querying the CQI mapping
table for UE categories (unit: dB).
For example, if the UE category is 8, is set to 6dB, and the adjusted CQI is 27, the
following information is obtained by querying the CQI mapping table for UE categories:
TBSIZE: 14411
number of HS-PDSCHs: 10
16
modulation mode: 16QAM
is -2dB
If PCPICH is 33 dBm, the HS-PDSCH power is 37 dBm (33 + 6 - 2).

If the number of available HS-PDSCH channelization codes in a cell is not less than 10,
the available power is not less than 37 dBm, and the UE data volume to be scheduled is
larger than 14411, then, TBSIZE is 14411, the number of HS-PDSCH channelization
codes is 10, the modulation mode is 16QAM, the power is 37dBm, and the data to be
scheduled of the UE is transmitted.
The above example is supposed in an ideal situation. In practice, code, power and data
volume may be insufficient for the CQI mapping table requirement. On the other hand,
UEs are scheduled in a descending order of relative priorities in one TTI. The UEs with
higher relative priorities are guaranteed to have sufficient resources, and the code and
power resources in a cell are used up as much as possible when the system selects the
number of HS-PDSCH channelization codes and power for a UE.
As shown in the following figure, suppose the UE categories, configurations and CQIs
of two UEs are identical, for example: UE category: 8; : 6 dB; adjusted CQI: 27, and the
to-be-scheduled data volume is sufficient. The number of HS-PDSCH channelization
codes is 15, and the available power of the cell is 6 W. 10 HS-PDSCH channelization
codes and 5 W power are used for scheduling of the first UE, and the remaining resource
includes five HS-PDSCH channelization codes and 1 W power. In this case, the available
TBSIZE for the second UE decreases to 4664.
Figure 3-1
Resource Allocation During UE Scheduling
HS-PDSCH 15
POWER 6W
HS-PDSCH 10
16QAM
TBSIZE 14411
POWER 37Dbm(5W)
HS-PDSCH 5
POWER 1W
HS-PDSCH 5
16QAM
TBSIZE 4664
POWER 30Dbm(1W)
The detailed process is described as follows:

The extended CQI mapping table is called CQI mapping extension table. The CQI
mapping extension table contains the number of channelization codes, modulation mode,
TBSIZE, and Es/No. For example, suppose the number of channelization codes is 10
17
and modulation mode is 16QAM, then the corresponding TBSIZE and Es/No table
entries are listed in the following table.
TBSIZE
Es/No
6554
14.56
7041
15.00
7430
15.36
7840
15.62
8272
15.89
8729
16.19
9047
16.39
9546
16.71
9894
16.94
10440
17.42
11017
17.93
11625
18.46
12048
18.78
12713
19.26
13177
19.60
13904
20.12
14411
20.50
Assume that the power meets the requirement during UE scheduling, and Es/No is
calculated in accordance with the following formula:
Es/No = -4.5 + CQI +
If CQI is 27, then the value of Es/No is 20.5. TBSIZE, which is 14411, is obtained by
querying the CQI mapping extension table based on the number of channelization codes
(10) and modulation mode (16QAM). The power required in this case is 37 dBm.
If the power is insufficient, for example, if it is only 36 dBm, Es/No needs to be adjusted
to (20.5 1), that is, 19.5. Suppose the number of channelization codes is 10, TBSIZE
and Es/No, which are 12713 and 19.26 respectively, are obtained by querying the CQI
mapping extension table. Then, the HS-PDSCH power is calculated in accordance with
the following formula:
18
HS-PDSCH power = 36 (19.5 19.26) = 35.76 dBm

If the number of available channelization codes is insufficient, the corresponding TBSIZE
and Es/No can be obtained by querying the CQI mapping extension table based on the
number of available channelization codes and UE Es/No. Then, the HS-PDSCH power is
calculated in accordance with Es/No.
If the to-be-scheduled data volume of UE is less than the selected TBSIZE, continue
querying the CQI mapping extension table and lower TBSIZE until it is slightly larger than
the to-be-scheduled data volume of the UE according to the HS-PDSCH power (Es/No),
the number of channelization codes and the modulation by the above steps.
The principle of querying the CQI mapping extension table is described as follows:
Firstly, based on the fixed modulation and fixed number of channelization codes, query
for TBSIZE through decreasing the HS-PDSCH power. If there is no suitable TBSIZE,
then continue querying the CQI mapping extension table after decreasing the number of
channelization codes. If there is no suitable TBSIZE yet, continue querying the CQI
mapping extension table after decreasing the modulation mode, for example, decrease
64QAM to 16QAM, or decrease 16QAM to QPSK.
Note: If is less than 0 (for example, UE category: 8; CQI > 25), to meet the testing
standard that the download rate is 85% of the air interface rate (namely, BLER is less
than 10%), the HS-PDSCH power will remain unchanged (namely, PCPICH + ). This is a
non-standard solution.
3.2
3.2.1
HSDPA Flow Control Algorithm

The HSDPA scheduler is used to implement the downlink packet forwarding function,
that is, receiving UE data over the lub interface and then transmitting the data through
the air interface.
The HS-PDSCH is shared among multiple UEs in a cell, and the total number of UEs and
channel quality of each UE keep changing, causing that the transmission rate of each UE
19
over the air interface keeps changing. Therefore, a flow control mechanism is required to
control the UE packet receiving rate over the lub interface to balance the inflow and
outflow of packets in the scheduler.
3.2.1.1
Flow Control Implementation Method

Flow control is implemented through the Frame Protocol (FP) on the user plane over the
lub interface. It uses the downlink HS-DSCH Capacity Request Frame (CRF) and uplink
HS-DSCH Capacity Allocation Frame (CAF).
The RNC will send a CRF to the Node B, when necessary, to trigger the flow control.
Node B will also send a CAF to the RNC, when necessary, to control the UE packet
inflow rate. For the formats of CAF and CRF, refer to the TS 25.435 protocol.
There are three trigger mechanisms available for flow control, namely, flow control will be
triggered in one of the following cases:
The Node B receives a CRF from the RNC.
The number of stacked packets on the UE side is larger than the upper limit or is
smaller than the lower limit.
The UE flow control timer expires.
In the first mechanism, it is the RNC that initiates the flow control. In the second
mechanism and the third mechanism, it is the Node B that initiates the flow control.
After the flow control is triggered, the RNC data receiving rate of the UE(Vin) is estimated
in accordance with the air interface rate of the UE, then the parameters related to the
CAF are calculated in accordance with Vin, finally, a CAF is sent. The following figure
describes the flow control procedure.
20
Figure 3-2
Flow Control Procedure
Flow Control
Trigger Event
Caculate Vin
Get credits interval

repetition period by Vin
Send Capacity
Allocation Frame
End
The minimum interval of UE flow control is set to 60 ms to avoid frequent trigger of flow
control.
3.2.1.2
Flow Control Algorithm

As mentioned above, flow control aims at maintaining a balance between UE packet
input and output. Furthermore, flow control also needs to assure there are sufficient
packets during UE scheduling so that the transmitting capability of UE air interface can
be fully made use of. The following describes the flow control algorithm:
Suppose each packet buffer area of UE is a leaky bucket. As shown in the following
figure,
V in
represents the inflow rate over the lub interface of every UE, corresponding
to the RNC data transmission rate.
Vout represents the transmission rate (outflow rate)
over the Uu interface of every UE. The target height of the leaky bucket is calculated in
21
accordance with the formula Haim Vout* BufferTime. BufferTime represents the
buffer time. The larger the number of to-be-scheduled UEs, the less outflow rate
Figure 3-3
Vout ..
Leaky Bucket Flow Control Principle
RNC
V in
Node B Capacity
Allocation
H AIM
H DIFFER = H AIM H CURRENT
PDU BUFFER
H DIFFER
Vin F (V , H
, H AIM )
out
DIFFER
H CURRENT
Vout
If
V in
>
Vout , the bucket height increases; if V in = Vout , the bucket height remains
the same; if
Take 80% of
V in
<
Vout , the bucket height decreases.
Haim as a median, the Node B does not proactively initiate flow control
when the inflow rate is between the upper and lower thresholds of this median. If
V in
is
larger than the upper threshold, which is usually set to 90% of the bucket height, the
V in
Node B initiates flow control to decrease the inflow rate
. If
the inflow rate is smaller
than the lower threshold, which is usually set to 70% of the bucket height, the Node B
also initiates flow control to increase the inflow rate
V in
Note: The BufferTime of the UE leaky bucket is constant, and is set to 150 ms by taking
into account various factors. The HSDPA scheduler offers a buffer area with a constant
22
size of not less than 150 Mbit for sharing among all UEs. Suppose BufferTime is 150 ms,
the buffer area far exceeds the current air interface capability requirement (1000 Mbit/s).
Because all UEs share the air interface capability in a cell, whether there are one or
several UEs in the cell does not make much of a difference for the required buffer area.
Because the value of BufferTime is constant, the higher the air interface rate of UE, the
more the buffer data packets used by UE.
In addition, the following two aspects are implemented in HSDPA packet scheduling to
cooperate with the HSDPA flow control algorithm:
According to characteristics of different services, the RNC configures and sends two
parameters to the Node B: DISCARDTIMERPRE and DISCARDTIMER. The Node B
employs the two parameters in different services scheduling processes. If a data
packet cannot be scheduled by the Node B in the specific UE data buffer in the
predefined time duration, the data packet will be discarded to relieve the congestion
in the data buffer of the Node B. The DISCARDTIMER parameter is not configured
for services that need accurate data transmission, such as interactive and
background class services. But for services that have high time-delay requirement
such
as
streaming
and
conversation
(VoIP
over
HSDPA)
class,
the
DISCARDTIMER parameter is not configured for them to discard data packets that
cannot be scheduled in time. The DISCARDTIMER time duration may be different,
for example, 4 s for streaming class services and 60-80 ms for conversation class
services with higher time-delay QOS requirement.
Introducing sliding window management and T1 timer mechanism to stop related

MAC-hs PDU retransmission which cannot be accepted by the UE even after
multiple retransmissions.
Node B sliding window management is an algorithm for MAC-hs PDU transmission

management in the scheduling process. It aims for
preventing a UE from receiving
MAC-hs PDUs with unclear TSNs. The Node B scheduling algorithm employs T1 timer to
stop MAC-hs PDU retransmission that cannot be accepted by UE even after multiple
retransmissions.
T1 timer(T1) and the window size parameter (MACHSWINSIZE) of MAC-hs transmission
and receiving side are delivered to the Node B and UE through signaling respectively.
23
Sliding window management refers to waiting MAC-hs PDU at the lower edge of the
window to be received accurately. When the number of times of retransmission exceeds
the threshold or the T1 timer expires, the window will slide forward.
TrsWindow_LowerEdge is the lower edge of the window,
Suppose
TrsWindow_LowerEdge
represents the minimum TSN of the MAC-hs PDU which is waiting for ACK response. If
ACK response is received by the MAC-hs PDU successfully or the MAC-hs PDU is
discarded due to the number of times of retransmission exceeding the threshold or the
T1 timer expiring, TrsWindow_LowerEdge slides to the next minimum TSN of the
MAC-hs PDU which is waiting for ACK response. The MAC-hs PDU with smaller TSN
than TrsWindow_LowerEdge will not be retransmitted, and the MAC-hs PDU with
bigger TSN than TrsWindow_LowerEdge + MACHSWINSIZE cannot be transmitted
(namely, waiting to be transmitted) .
For example, MACHSWINSIZE is 12, TrsWindow_LowerEdge is 0, see the following
figure,
TSN=12 MAC-hs PDUs outside of the window are not allowed to be transmitted.
If NACK response is received for TSN=0 MAC-hs PDUs and the PDUs are discarded due
to the number of times of retransmission exceeding the threshold and T1 timer expiring,
and if ACK response
is received for TSN=1 MAC-hs PDUs and NACK response is
received for TSN=2 MAC-hs PDUs, TrsWindow_LowerEdge will slide to 2. In this case,
MAC-hs PDUs (TSN=12,13) are allowed to be transmitted, but MAC-hs PDUs (TSN=14)
are still waiting to be transmitted, and MAC-hs PDUs (TSN<2) are not allowed to be
retransmitted.
The default configuration for MACHSWINSIZE is 16. For MIMO dual stream transmission
services and DC-HSDPA services, MACHSWINSIZE is equal to MIN (by default,
configured to be MACHSWINSIZE*2, 32).
10
11
12
13
14
TRANSMIT_WINDOW_SIZE
TRANSMIT_WINDOW_SIZE
The Node B employs the T1 timer to stop retransmission of related MAC-hs PDUs.
24
At the UE side, if no timer T1 is active, timer T1 is started when an MAC-hs PDU with
TSN > next_expected_TSN is correctly received, in other words, the UE will start timer
T1 based on the next TSN MAC-hs PDU successfully received. But the Node B does not
know accurately whether the UE receives the next TSN MAC-hs PDU or not, thus the
Node B will start timer T1 after receiving ACK response for the next TSN MAC-hs PDU
from the UE. ACK response delay or ACK loss may cause the Node B to start timer T1
later than the UE, in other words, the Node B will properly postpone the PDU discarding.
However, the mechanism of retransmission priority and retransmission number limitation
ensures that the Node B can discard specific MAC-hs PDUs in time and will not block the
next TSN PDU transmission.
The interval of timer T1 ranges from 10 ms to 400ms, and the default value is 50 ms at
the RNC.
3.2.2
Measurement of HS-DSCH Required Power

According to the TS25.433 protocol, there are three types of common measurement:
Transmitted carrier power of all codes not used for HS-PDSCH or HS-SCCH
transmission
HS-DSCH required power
HS-DSCH provided bit rate
Among these common measurements, HS-DSCH required power is a key reference

parameter of RNC admission control. The following gives a detailed description of
HS-DSCH required power.
According to the TS25.433 protocol, the HS-DSCH required power value indicates the
minimum necessary power for a given priority class to meet the GBR for all UEs
belonging to this priority class. It is expressed in permillage of the cell maximum
transmission power.
The basic principle is to pre-calculate required power for GBR rate with 100 ms time
duration according to the recent historical power and throughput rate.
25
The Node B can schedule multiple packet data QUEs for every UE. For one
scheduled packet data QUE, the required power is the sum of HS-SCCH required
power and HS-DSCH required power. HS-SCCH required power can be calculated
by HS-SCCH reserved power/QUE number. The HS-DSCH required power
calculation process is described as follows:
Get CQI according to GBR.
GbrTbSize = GBR/500, check the TS25.214 CQI table and get GbrCqi
according to GbrTbSize.
Statistics of scheduled packet data QUE every 100 ms.

HS-DSCH accumulated power: AccuDschPwr
Accumulated transmitted BITNumber: AccuSchBitNum
According the actual scheduling, calculate the actual corresponding CQI:

SchTbSize
= AccuSchBitNum / SchNum, thereinto, SchNum is the
scheduled
times in 100 ms, check the TS25.214 CQI table, get SchCqi according to
SchTbSize.
Calculate HS-DSCH required power.

The actual average HS-DSCH power is: SchDschPwr
= AccuDschPwer /
SchNum
HS-DSCH required power = SchDschPwr - (SchCqi - GbrCqi)
At the same time, pay attention that HS-DSCH required power cannot be
higher than the pilot power+MPO, and cannot be lower than the pilot power
-10dB.
Filter HS-DSCH required power and report it to the RNC.

The current reported HS-DSCH required power = last time reported HS-DSCH
required power * (1- gdfMeasFilterCoeff) + current calculated HS-DSCH
required
power
gdfMeasFilterCoeff,
thereinto
the
filtering
factor
gdfMeasFilterCoeff = 0.015625.
26
3.3
Impact of HSPA+ on the Scheduler
3.3.1
Impact of 64QAM Modulation on the Scheduler

After introducing the HSDPA and HSUPA technologies respectively in R5 and R6
protocol versions, the 3GPP organization now introduces new technologies in R7 and
subsequent versions to enhance HSPA performance. After the HSDPA technology is
introduced to the WCDMA system, the MAC-ehs layer is added to both the Node B and
UE to be compatible with functions of the original MAC-hs layer. For the HSDPA
scheduler of Node B, R7 is primarily characterized by two technologies: 64QAM and
enhanced L2.
Introducing of the HSPA+ technology does not have any impact on the HSDPA
scheduling
and
flow
control
algorithms.
But
it
has
certain
impact
on
HS-SCCH/HS-PDSCH code selection, and packetization of MAC-ehs PDU.

After the 64QAM modulation technology is used, 64QAM improves the modulation
efficiency by 50%, and accordingly, the peak rate of a single UE increases by 50% and
reaches 21.6 Mbps compared with HS-DSCH of R5.
The major impact of 64QAM on the HSDPA scheduler lies in the change of modulation
mode information carried on HS-SCCH and HS-PDSCH code-set information.
In the R5 TS25.212 protocol, the modulation mode information and HS-PDSCH code-set
information carried on HS-SCCH are defined as follows:
xccs,1, xccs,2, xccs,3 = min(P-1,15-P)
refers
to
the
number
of
HS-PDSCH
channelization codes.
xccs,4, xccs,5, xccs,6, xccs,7 = |O-1-P/8 *15|
O refers to the starting code number of
HS-PDSCH.
ms,1
0 if QPSK
1 if 16QAM
After 64QAM modulation mode is introduced, if xms,1 is 0, the modulation mode is QPSK;
if xms,1 is 1, the modulation mode is 16QAM or 64QAM.
27
0 if QPSK
ms ,1
1 otherwise
To differentiate 16QAM from 64QAM, the lowest BIT xccs,7 in HS-PDSCH code
information bits (xccs,4, xccs,5, xccs,6, xccs,7) is used. Specifically, if xccs,7 is 0, the modulation
mode is 16QAM; if xccs,7 is 1, the modulation mode is 64QAM.
0 if
xccs,7 1 if
16QAM
64QAM
The lowest BIT position (xccs,7 in R5) of |O-1-P/8 *15| is subject to the even/odd position
of HS-SCCH code demodulated by the UE. If the HS-SCCH code position is 1, or 3, xccs,7
is 1; if the HS-SCCH code position is 2, or 4, xccs,7 is 0. In other words, the following
equation must be met:
|O-1-P/8 *15| mod 2 = (HS-SCCH number) mod 2
If 16QAM or 64QAM modulation mode is selected, the available HS-PDSCH start code
numbers are listed as follows:
Table 3-2
HS-PDSCH Starting Code Numbers

Odd Position of HS-SCCH Code
P8
13
P<8
24
8 10
Even Position of HS-SCCH Code

246
12 14
13
7 9
11
13
15
To break the downlink data transmission rate bottleneck caused by fixed AM RLC PDU
length and RLC transmit window size defined in R6 and earlier versions, the enhanced
L2 technology enables the RLC layer of the RNC to support variable RLC PDU lengths.
The corresponding MAC-ehs SDU length changes, and the length field L in the MAC-ehs
PDU is used to indicate MAC-ehs SDU length.
The multiplexing of logical channel data shifts from the RNC to the Node B, that is,
several dedicated logical channels are multiplexed into one MAC-d stream and are
distinguished through Logical Channel IDs (LCH-IDs) in MAC-ehs PDU. When MAC-ehs
is used, LCH-ID multiplexing replaces C/T multiplexing in MAC-d. The HSDPA scheduler
receives and stores the LCH-ID field from the HS-DSCH DATA FRAME TYPE2, and
forwards it through the MAC-ehs PDU. For the frame format of HS-DSCH DATA FRAME
TYPE2, refer to the TS25.435 protocol.
28
MAC-ehs SDUs are of variable lengths. To avoid failure in transmitting a complete

MAC-ehs SDU because the length of MAC-ehs SDU exceeds that of the MAC-ehs PDU,
the MAC-ehs PDU has a Segmentation Indication (SI) field added to support segmented
transmission of MAC-ehs SDUs.
For the format of MAC-ehs PDU, refer to the TS25.321 protocol.
If both the Node B and UE support enhanced L2, the RNC determines the type of
signaling and service, that is, fixed or flexible PDU size, based on the settings of the
parameter RlcSizeSuptType. This parameter is a switch used to indicate RLC PDU
size type configured for signaling and services. It has four values:
1: Signalling Fixed Mode, Service Flexible Mode
2: Signalling and Conversational service Fixed Mode, Service Flexible Mode
3: All Flexible Mode
4: All Fixed Mode
The RLC information reassignment is required when a UE performs handover between

cells that support flexible size and those that do not support flexible size, or when the
transport channel type switches between HS-DSCH and DCH or FACH. In the case of LI
size inconsistency after the handover or switching, the RESET or MRW procedure must
be initiated to re-establish RLC. To avoid signaling RB RLC reestablishment, fixed PDU
Size is recommended for signaling RB, and the corresponding RLC PDU size is set to
144 bits and LI size is constantly 7 bits. After new technologies including 64QAM are
used, flexible PDU size must be used for service RB, thus the parameter
RlcSizeSuptType is set to 1 by default.
The maximum length of Data field in the RLC PDU is 1504 bytes (including the lengths of
SN and LI) as stipulated in the protocol. The actual maximum available SDU size is
configured by the RNC through the parameter CMaxPDUSize and NonCMaxPDUSize.
CMaxPDUSize is configured for conversational service and NonCMaxPDUSize is
configured for non conversational service. In view of protocol header overhead at bottom
layers (including the FP layer and IP/UDP layer) during data transmission,
NonCMaxPDUSize is set to 600 bytes (4800 bits) by default to avoid efficiency decrease
due to segmentation at the transmission layer.
29
3.3.2
Impact of MIMO on the Scheduler

The impact of introducing MIMO in R7 on the scheduler includes the following aspects:
The PF scheduling algorithm needs to be modified for MIMO UEs.
The MIMO single/dual stream selecting algorithm must be added before the
scheduling priority calculation. The MIMO single/dual stream selecting algorithm is
to preliminarily choose single or dual stream transmission for the MIMO UE in the
current TTI.
Impact on the TFRC selection algorithm.
VAM technology.
Enhanced-VAM (EVAM) technology.
The following describes the impacts in detail:

1.
PF scheduling algorithm needs to be modified according to MIMO UE

Refer to section 3.1.2.3 for the description related to MIMO UE.
2.
MIMO single/dual stream selecting algorithm

p ref
p ref
The UE decides preferred precoding control indication (PCI) vector matrix w1 , w2
combining with CQI and transmits to the Node B through uplink HS-DPCCH signaling.
Based on the PCI/CQI report, the Node B packet scheduling module decides and sends
to the UE with single or dual stream transmission mode, TBSize, modulation mode in the
next TTI. At the same time, the Node B informs the UE of the precoding weight w2 which
is put in precoding weight indication bit in the first part of downlink HS-SCCH subframe.
The PCI and CQI compositive coding formats are provided when the UE is configured to
be in MIMO working mode. The UE must support two types of CQI reports: Type A CQI
report and Type B CQI report. For Type A CQI report and Type B CQI report, the CQI
value range is (030) for single stream transmission. For Type A CQI report, the CQI
value range is (0255) for dual stream transmission.
Type A CQI report is a CQI report format which the UE decides transmission block
number (one block or two blocks) according to the current channel condition. When one
transmission block is selected by the UE, the preferred primary precoding vector (PCI)
30
from HS-DPCCH will be used in the Node B to precode the primary transmission block.
When two transmission blocks are selected by the UE, the preferred primary precoding
vector and orthogonal precoding vector (PCI) will be used in the Node B to precode the
primary and secondary transmission blocks. Type A CQI report may include one or two
transmission formats according to the number of transmission blocks.
Type B CQI report is a CQI report format which the UE decides single transmission block
according to the current channel condition. The preferred primary precoding vector (PCI)
from HS-DPCCH is used in the Node B to precode the primary transmission block, and
there is no secondary transmission block.
For the two different types of CQI reports (Type A and Type B), the UE use the following
method to achieve different report rates configured by the RAN:
The report rate for Type A CQI report is N_cqi_typeA/M_cqi of the reported available CQI
rate, the remaining rate (M_cqi-N_cqi_typeA) is used for Type B CQI report. The MIMO
CQI report rate that can be used is the same as SISO, which is the function of CQI
feedback duration k and CQI repeating factor N_cqi_transmit.
According to TS25.214v7.9.0, when the following formula is tenable:
5 CFN m 256chip 7680chip
mod M_cqi N_cqi_typeA k k (2ms)

with
,
the UE will send Type A CQI report; otherwise, send Type B CQI report.
When MIMO Activation Indicator in HS-DSCH FDD Information signaling is sent by the
RNC, the Node B will feed back MIMO N/M Ratio (N_cqi_typeA/M_cqi) to the RNC in
HS-DSCH Information Response signaling. The RNC sends MIMO N/M Ratio to the UE
through RRC signaling. The UE sends type A and type B CQI reports according to MIMO
N/M Ratio, the Node B receives type A and type B CQI reports according to MIMO N/M
Ratio. The MIMO N/M Ratio value can be 1/2, 2/3, 3/4, 4/5, 5/6, 6/7, 7/8, 8/9, 9/10, 1/1. At
present, MIMO N/M Ratio can be set to 1/2 or 1, and the default value is 1, which
indicates that the UE always reports Type A CQI report.
The MIMO single/dual stream selecting algorithm flow is described as follows:
After outer loop process of the UE reported CQI, the Node B determines whether the UE
CQI report is Type A or Type B. If it is Type A, go to step 2; if it is Type B, go to step 3.
31
For Type A CQI, if its single stream CQI, the Node B employs single stream scheduling
in the current TTI; if its dual stream CQI, the Node B employs dual streams scheduling in
the current TTI;
For Type B CQI, the Node B checks whether the recent historical type A CQI report is
single stream or dual streams before the current TTI. If its single stream, the Node B
employs single stream scheduling for the UE in the current TTI based on Type B CQI
report. If its dual stream, the current TTI single stream TBSize and the sum of recent
historical dual streams TBSize are compared, if the current TTI single stream TBSize is
larger, the Node B employs single stream scheduling for the UE in the current TTI based
on Type B CQI report; otherwise, the Node B employs dual streams scheduling based on
Type A CQI report;
Under small data buffer conditions, the dual streams process is changed to single stream
process.
3.
Impact on the TFRC selection algorithm

The TFRC selection algorithm for MIMO UE single stream scheduling is the same as that
for traditional UEs in section 3.1.3. But there is some difference for MIMO UE double
stream scheduling in the TFRC selection algorithm:
Firstly, modify the CQI of MIMO primary and secondary stream individually with the same
principle described in section 3.1.3.2.
According to the modified CQI of primary and secondary stream, without consideration of
code and power resource, select the TBSIZE of the two streams, calculate the needed
power, and calculate the power per code channel and code rate.
According to the cell available power, calculate the number of available UE code
channels based on the principle of no change to power of per code channel
According to the number of available code channels, select the TBSIZE based on the
principle of no change to code rate
Modify TBSIZE according to the volume of packet data to be scheduled.
4.
VAM technology
The primary and secondary pilot configuration and Virtual Antenna Mapping (VAM)
technology are deployed when MIMO UEs and traditional non-MIMO UEs are supported
in hybrid networking. The VAM technology is used for power balance between two PAs
32
and load balance between two transmission channels. For MIMO UEs, under the single
1 j
w2pref
2
stream scheduling condition, PCI codebook restrictions of
1 j
2 on
the UE side is needed to implement power balance between two PAs. In order to reduce
the impact of the secondary pilot on traditional non-MIMO UEs, the secondary pilot
power can be configured as half of the primary pilot, but few MIMO commercial terminals
support the configuration.
5.
EVAM technology
The EVAM technology is introduced into HSPA+ after the VAM technology. getting the
best SINR of signals received by traditional non-MIMO UEs. Through adjusting the VAM
matrix, it gets the best SINR of signals received by traditional non-MIMO UEs, and
improves the CQI of traditional non-MIMO UEs without affecting MIMO UEs.
It
increases the throughput of non-MIMO UEs, improves the network fairness, and
improves the performance of networks with fewer MIMO UEs.
3.3.3
Impact of DC-HSDPA on the Scheduler

After introducing DC-HSDPA in R8, the PF scheduling algorithm is used for Dual-Carrier
(DC) UEs, and the scheduling relative priority formula for DC UEs needs to be adjusted
accordingly. Refer to section 3.1.2.3 for details related to the DC UE.
The queuing and scheduling rule according to UE relative scheduling priorities:
On carrier 1, single-carrier UEs and DC UEs are queued and scheduled according to
their relative priorities.
On carrier 2, single-carrier UEs and DC UEs are queued and scheduled according to
their relative priorities.
3.3.4
Impact of Enhanced CELL-FACH on the Scheduler

The enchanced CELL_FACH states include two cases: common E-DCH and RACH on
uplink. If common E-DCH is configured, the UE reports feedback information such as
CQI and ACK/NACK through HS-DPCCH, and the NodeB performs scheduling
operations according to the reported CQI. However, if the UE is not configured to be in
33
common E-DCH state, it reports the measurement report through RACH. Because the
measurement report contains CPICH RSCP, Ec/Io or path loss, the Node B cannot use
the measurement report to directly perform scheduling operations. In this case, the
measurement report must be converted to CQI form before scheduling operations are
performed.
Suppose the UE measurement report contains CPICH Ec/Io, the relevant conversion
method is given as below:
CQI ceil(Ec Io 10log16 MPO 4.5)
3.3.5
Impact of CPC on the Scheduler

The CPC affects the HSDPA scheduler in the following two aspects:
First, in which situation does HSDPA scheduler send CPC activation/deactivation order?
In other words, what triggers HS-SCCH orders to activate/deactivate DTX, DRX or
HS-SCCH Less operation?
Second, how does the HSDPA scheduler determine order sending priority?
As for the first question, different trigger scheme for different vendors.
For the second question, the order sending priority is determined on the basis of the
following rules.
Order has a higher priority than data-type HS-SCCH.
If an order is used to change a serving cell, it has the highest priority. Below it is the
order used to deactivate DTX/DRX, followed by the orders of HS-SCCH-less and
activation DTX/DRX operations. The lowest priority is assigned to an order used to
activate/deactivate HS-SCCH in a secondary serving cell.
If two orders have the same priority, the retransmission order has higher priority
than the first transmission order. If two orders have different priorities, the
low-priority order always has lower priority than the high-priority order regardless of
retransmission or first transmission.
34
3.3.6
Impact of CS Voice Service over HSPA on the Scheduler

CS voice services belong to conversational services. It has strict requirement for delay.
Therefore, the PF scheduling algorithm must be changed after CS Voice Service over
HSPA is introduced. The algorithm must add a delay weight factor to ensure quality and
capacity of CS voice services. Refer to section 3.1.2.3 for details.
3.3.7
Impact of DC+MIMO+64AM on the Scheduler

The MIMO+64QAM technology is introduced in R7 to support high-order modulation..
Then
the
technology
which
combines
DC
with
MIMO+64QAM
is
called
DC+MIMO+64QAM technology is introduced. DC+MIMO+64QAM takes advantage of

these three technologies, and it increases data rate of UEs not only in the centre of but
also at the edge of a cell, which enables users to experience the improvement of network
performance. Refer to section 3.3.1,3.3.2 and 3.3.3 for details.
3.3.8
Impact of Dual-Band DC-HSDPA on the Scheduler

According to the R9 protocol, DB DC-HSDPA allows two carriers to be located in
different frequency bands. Except for
frequency combination, the DB DC-HSDPA and
DC HSDPA functions are similar. Refer to section 3.3.3 for details.
3.3.9
Impact of 4C-HSDPA on the Scheduler

This feature further extends the application of DC-HSDPA in 3GPP Rel8 and DB
DC-HSDPA in Rel9. It allows that four carriers are concurrently used for an HSDPA user
and increases the theoretical peak rate in downlink for a single user to 84 Mbps via the
combination of four carriers and 64QAM. Refer to section 3.1.2.3 for details.
The queuing and scheduling rule according UE relative scheduling priority:
On carrier 1, single-carrier UEs and MC UEs are queued and scheduled according to the
relative priorities.
On carrier 2, single-carrier UEs and MC UEs are queued and scheduled according to the
relative priorities.
35
On carrier m, single-carrier UEs and MC UEs are queued and scheduled according to
the relative priorities.
3.3.10
Impact of DB DC-HSDPA+MIMO on the Scheduler

This feature is an improvement for DB DC-HSDPA in 3GPP Rel. 9. It allows that the
MIMO can be concurrently used with DB DC-HSDPA. With this feature, the DB
DC-HSDPA user peak rate can be increased. Please see section 3.3.1,3.3.2 and 3.3.8
for details.
3.3.11
Impact of 4C-HSDPA+MIMO on the Scheduler

4C-HSDPA is introduced in R10. The technology combines 4C with MIMO.
4C+MIMO+64QAM takes advantage of these three technologies, and the user peak rate
in downlink in theory can reach 168 Mbps. Refer to section 3.3.1,3.3.2 and 3.3.9 for
details.
Parameters
4.1

Table 4-1
HSDPA Fast Scheduling OMMB Parameters List

Recom
Managed
Object
Logic Name
Value
Parameter Description
Range
Unit
Default mende
Value
d
Value
NodeBFun
ction
NodeBFun
ction
NodeBFun
ction
HSDPASPI SPI
hsdpaSched
uleAlg
weight
[1~2000] None
long:0:R
hsdpaSchedule algorithm
hsdpaPFWgt hsdpaPF weight
R,1:MAX None
C/I,2:PF
long:[1~6
]
None
36
Table 4-2
HSDPA Fast Scheduling OMMR Parameters List

Recom
Managed
Object
Logic Name
Value
Parameter Description
Range
Unit
Default mende
Value
d
Value
UHspa
UMPO
UMPO
HsdschTotP HSPA Total Downlink Power

wrMeth
Allocation Method
MeasPwrOff HS-PDSCH
set
app64QamI
nd
Power Offset
Actual
Configuration
None
[-6,13]
dB
none
none
none
none
600
600
65
65
of
HSDPA Evolution Function [0,1]

of UE in the Cell(64QAM)
Actual
UMPO
Measurement
[1,2]
Configuration
of
appMimoInd HSDPA Evolution Function [0,1]

of UE in the Cell(MIMO)
Actual
UMPO
Configuration
of
appDcHsdp HSDPA Evolution Function

aInd
of
UE
in
the
[0,1]
Cell(DC-HSDPA)
UHspa
RlcSizeSupt
Type
Support Type of Downlink

RLC
Flexible
PDU
Size [1 ,2,3]
Format
Downlink
URncFunct NonCMaxPd Non-Conversational Service

ion
uSize
Maximum MAC-d PDU Size
[1~1505]; byte
Extended(bytes)
URncFunct CMaxPduSi
ion
UCHspa
ze
HSNBASSI
ND
Downlink
Conversational
Service Maximum MAC-d [1~1505] Byte

PDU Size Extended(bytes)
HS-PDSCH Code NodeB
Assignment
Support [0,1]
none
Indicator
37
4.2

None
Related Counters and Alarms
5.1
Related Counters
5.1.1

Table 5-1
HSDPA Fast Scheduling Counter List
Counter ID
Name
C372490000
Maximum power reserved for HSDPA
C372490001
HSDPA Maximum power
C372490002
HSDPA Minimum power
C372490003
HSDPA Average power
C372490004
HSDPA DL Power distribution (-inf, 31dBm]
C372490005
HSDPA DL Power distribution (31dbm, 32dBm]
C372490006
C372490007
C372490008
C372490009
C372490010
C372490011
C372490012
C372490013
C372490014
C372490015
C372490016
C372490017
38
C372490018
C372490019
HSDPA DL Power distribution (45dbm, +inf)
C372490020
HSDPA Radio
C372490021
Maximum number of HS-PDSCHs Used for HSDPA
C372490022
Minimum number of HS-PDSCHs Used for HSDPA
C372490023
Average number of HS-PDSCHs Used for HSDPA
C372490024
Number of HS-PDSCHs Used for HSDPA
C372490025
Number of Reported CQI 0
C372490026
C372490027
C372490028
C372490029
C372490030
C372490031
C372490032
C372490033
C372490034
C372490035
C372490036
C372490037
C372490038
C372490039
C372490040
C372490041
C372490042
C372490043
C372490044
C372490045
C372490046
C372490047
C372490048
C372490049
Load - Downlink
39
C372490050
C372490051
C372490052
C372490053
C372490054
C372490055
C372490056
Average Value of Reported CQIs
C372490073
Number of Incorrect CQIs
C372490074
Number of CQI 0 Used for Scheduling
C372490075
C372490076
C372490077
C372490078
C372490079
C372490080
C372490081
C372490082
C372490083
C372490084
C372490085
C372490086
C372490087
C372490088
C372490089
C372490090
Number of CQI 16Used for Scheduling
C372490091
C372490092
C372490093
C372490094
C372490095
C372490096
C372490097
40
C372490098
C372490099
C372490100
C372490101
C372490102
C372490103
C372490104
C372490105
Average Value of CQIs Used for Scheduling
C372490122
HSDPA Traffic - QPSK
C372490123
HSDPA Traffic - 16 QAM
C372490124
HSDPA Traffic - 64 QAM
C372490125
HSDPA - QPSK Trafic Ratio
C372490126
HSDPA - 16 QAM Trafic Ratio
C372490127
HSDPA - 64 QAM Trafic Ratio
C372490128
HSDPA Duration
C372490129
HSDPA Average Throughput per user
C372490130
HSDPA Average user rate distribution 0 - 8 Kbits/s
C372490131
HSDPA Average user rate distribution 8 - 32Kbits/s
C372490132
C372490133
C372490134
C372490135
C372490136
C372490137
C372490138
C372490139
HSDPA Average user rate distribution 1024 1280Kbits/s
C372490140
C372490141
C372490142
41
C372490143
C372490144
C372490145
C372490146
C372490147
C372490148
C372490149
C372490150
C372490151
C372490152
C372490153
HSDPA Average user rate distribution 7168Kbits/s - inf
C372490160
Number of Received NACK - HSDPA
C372490161
Number of Received ACK - HSDPA
C372490162
Number of Received DTX - HSDPA
C372490163
Rate of NACK and DTX - HSDPA
C372490164
Number of Residual Error Blocks - HSDPA
C372490165
Total size of successfully-sent MAC-HS PDU
C372490166
Number of HSDPA Scheduled UE
C372490167
Number of successfully-sent MAC-HS PDUs
C372490168
Number of Retransmitted MAC-HS PDUs
C372490169
HSDPA MAC-HS Efficiency
C372490170
Discarded MAC-D PDU rate
C372490171
HSDPA Scheduling Rate
C372490172
HSDPA RAB Activity Factor
42
C372490173
Number of Discarded Data Frames Due to HS-DSCH

Congestion at Iub
C372490174
Number of Capacity Allocation Frames due to Too Much

Data
C372490175
Number of Capacity Allocation Frames due to Too Little

Data
C372490176
HSDPA Scheduling Delay
C372490177
Throughput Rate of 64QAM
C372490178
Scheduling Ratio of 64QAM
C372490179
Two Cell Scheduling of DC User
C372490180
One Cell Scheduling of DC User
C372490190
Retransmission Ratio of 64QAM
C372490191
Retransmission Throughput of 64QAM
C372490192
Total size of Iub HSDPA MAC-d PDU
C372490193
Number of DC Secondary cell Deactive
C372490194
Number of HSDPA User
C372490198
Number of MIMO User
C372490199
Ratio of MIMO User
C372490200
Throughput Rate of MIMO User
C372490201
Number of 64QAM User
C372490202
Ratio of 64QAM User
C372490203
Throughput of MAC-EHS
C372490204
Scheduling of MAC-EHS
C372490205
Throughput Rate of MAC-EHS
C372490206
Number of DL_DRX User
C372490207
Number of UL_DTX User
C372490208
Ratio of DL_DRX User
C372490209
Ratio of UL_DTX User
C372490210
Scheduling of 64QAM
C372490211
Ratio of MIMO+16QAM
C372490212
Scheduling of MIMO+16QAM
C372490213
Ratio of MIMO+64QAM
C372490214
Scheduling of MIMO+64QAM
43
C372490215
Ratio of DC+64QAM
C372490216
Scheduling of DC+64QAM
C372490217
Ratio of DC Two Cell Scheduling
C372490218
Throughput Rate of DC
C372490219
Scheduling of MAX HS-SCCH Power
C372490220
Non-HSPA DL SF Code Max Utilization
C372490221
Non-HSPA DL SF Code Average Utilization
C372490222
Non-HSPA DL SF Code Min Utilization
C372490223
HSPA DL SF Code Max Utilization
C372490224
HSPA DL SF Code Average Utilization
C372490225
HSPA DL SF Code Min Utilization
C372490226
DL SF Code Max Utilization
C372490227
DL SF Code Average Utilization
C372490228
DL SF Code Min Utilization
C372490229
4C-HSDPA scheduled time
C372490230
C372490231
C372490232
SC-HSDPA scheduled time
C372490233
4C-HSDPA downlink confirmed data traffic
C372490234
4C-HSDPA scheduled time * number of users
C372490235
C372490236
C372490237
C372490238
C372490239
SC-HSDPA downlink confirmed data traffic
C372490240
SC-HSDPA scheduled time * number of users
C372490247
Number of TTIs for downlink DC+MIMO demodulation
C372490248
Number of TTIs for downlink DC+MIMO+64QAM

demodulation
C372490249
Number of TTIs for downlink DC
C372490250
Number of HSDPA packets once successfully

transmitted at physical layer
44
C372490251
Number of HSDPA packets twice successfully

C372490252
Number of HSDPA packets three times successfully

C372490253
Number of HSDPA packets four times successfully

C372490254
Number of HSDPA packets four times unsuccessfully

C372490255
Number of active DC-HSDPA users in a cell
C372490256
Sum of number of DC-HSDPA dual-stream users * time
C372490265
Number of TTIs in an MIMO+64QAM adopted cell
C372490266
Number of TTIs in an MIMO+16QAM adopted cell
C372490267
Number of TTIs in an MIMO+QPSK adopted cell
C372490268
Total number of HSDPA UE scheduling TTIs
C372490269
Total number of HSDPA active TTIs
C372490270
Total number of TTIs when single HS-SCCH power

reaches a peak
C372490271
How many times does HS-SCCH total power

consumption reach a peak
C372490272
Total number of TTIs when HS-SCCH is used up and

scheduled
C372490273
Total number of TTIs when HS-PDSCH has no available

power
C372490274
Number of TTIs when all HS-PDSCH channels have

been scheduled
C372490275
Total number of TTIs when all HS-PDSCH power has

been scheduled.
C372490276
Total number of TTIs when all HS-PDSCH channels

have been scheduled and all power has been scheduled
C372490277
Number of sectors in a cell
C372490278
Number of HS-PDSCH available SF16 codes
C372490000
Number of HS-PDSCH used SF16 codes
C372490001
Number of Discarded MAC-D PDU
C372490002
Number of TTI with scheduling data but not scheduled
C372490315
Times of 1 HS-PDSCH used for HSDPA
45
C372490316
Times of 2 HS-PDSCHs used for HSDPA
C372490317
C372490318
C372490319
C372490320
C372490321
C372490322
C372490323
C372490324
C372490325
C372490326
C372490327
C372490328
C372490329
C372490330
Times of 1 HSDPA user scheduled in one TTI
C372490331
Times of 2 HSDPA users scheduled in one TTI
C372490332
Times of 3 HSDPA users scheduled in one TTI
C372490333
Times of 4 or more HSDPA users scheduled in one TTI
C372490334
Times of power limit in HSDPA scheduler
C372490335
Times of code limit in HSDPA scheduler
46
5.1.2

None
5.2
Related Alarms
5.2.1

None
5.2.2

None
Engineering Guide
6.1
Application Scenario
With the wider coverage of the network, users are demanding higher download rate. But
the traditional R99 failed to meet the needs of users. However, the HSDPA technology
has greatly improved the downlink rate. Enable HSDPA fast scheduling and HSDPA flow
control to improve the user experience.
6.2
Feature Activation Procedure
6.2.1

In the configuration resource tree window, open the [Modify Area->Managed
Element->UMTS Logical Function Configuration->UTRAN Cell] interface and set the
parameter HSPA Support Method, as shown in Figure 6-1
47
Figure 6-1
Configure HSPA Support Method of the cell

parameter HSDPA Function Status, as shown in Figure 6-2
Figure 6-2
Configure HSDPA Function Status
48
6.2.2

parameter HSDPA Function Status, as shown in Figure 6-3.
Figure 6-3
Configure HSDPA Function Status

Element->UMTS Logical Function Configuration->UTRAN Cell->Hspa Configuration In A
Cell] interface and set the parameter
Number of HS-PDSCH, Minimum Number of
HS-PDSCH, Maximum Number of HS-PDSCH, as shown in Figure 6-4.
49
Figure 6-4
Configure the number of HS-PDSCH
6.3
Feature Validation Procedure
6.3.1

HSDPA Fast Scheduling
Test Item
1.
WCDMA system is ready.
2.
Cell1 supports HSUPA, HSDPA and DCH.
Preconditio
3.
UE1 camps on Cell1 in Idle mode.
ns
4.
UE1 supports R5. The HSDPA category is 8.
5.
UE1 subscribes interactive or background service, MBR=

UL2Mbps/ DL8Mbps.
Test Steps
Anticipative
Result
1.
UE1 activates a PS call in Cell1 and starts FTP downloading.
2.
Maintain the downloading for 90s.
3.
Deactivate the PDP.
1.
In step1, the PS call is established on DCH /HS-DSCH channel

successfully.
2.
In step2, the throughput of DL is stable.
50
6.3.2

Test Item
HSDPA Flow Control

1.
WCDMA system is ready.
2.
Cell1 supports HSDPA and DCH.
Preconditio
3.
Configure 10 HS-PDSCH channels supported in Cell1.
ns
4.
UE1 and UE2 camp on Cell1 in Idle mode.
5.
UE1 and UE2 support R5 and subscribe background service,

MBR= UL2Mbps/ DL8Mbps.
1.
UE1 activates a PS service on DCH/HS-DSCH in Cell1 and starts

FTP downloading.
2.
UE2 activates another PS service on DCH/HS-DSCH in Cell1 and

starts FTP downloading. The PS service of UE2 is established
Test Steps
successfully by downgrading the DL data rate of UE1.

3.
Deactivate the PDP of UE2. The data rate of UE1 is recovered to

the original value.
4.
Deactivate the PDP of UE1.
1.
In step1, the data rate of UE1 is more than 5.8M.
2.
The DL data rate of UE1 is downgraded after UE2 is admitted into

Cell1.
Anticipative
3.
Neither the DL data rate of UE1 nor UE2 can reach its maximum
data rate because of the shortage of HS-DSCH codes. Nodeb will
Result
limit the data rate that RNC sends through HSDPA flow control
function.
4.
The data rate of UE1 is recovered to the original value after the
PDP of UE2 is deactivated.
6.4
Feature Deactivation Procedure
6.4.1

Managed Object. logic name
GUI Name
Default
Deactivation
Value
Value
HSDPA
UUtranCellFDD.hsdStat
Function
1: Active
0: Inactive
Status
51
6.4.2

Managed Object. logic
name
HSDPA Function
UUtranCellFDD.hsdStat
Default Value
GUI Name
Status
Deactivation
Value
1: Active
6.5
Impact on the Network
6.5.1
0: Inactive
None
6.5.2

None
Abbreviation
Abbreviation
Full Name
rd
3GPP
3 Generation Partnership Project
64QAM
64 grade Quadrature Amplitude Modulation
AGPS
Assisted GPS
AMR
Adaptive Multi Rate
BB
Base Band
BBU
Base Band processing Unit
BCCH
Broadcast Control Channel
BITS
Building Integrated Timing Supply
BPK
Base band Processing type K for UMTS
BSC
Base Station Controller
BSP
Board Support Package
BTS
Base Transceiver Station
CAPEX
CAPital EXpenditure
52
CC
Control & Clock
CE
Channel Element
CN
Core Network
CORBA
Common Object Request Broker Architecture
CPRI
Common Public Radio Interface
CS
Circuit Switch
CS
Carrier Sector
DL
Down Link
DTX
Discontinuous transmission
EDGE
Enhanced Data rates for GSM Evolution
E-EDGE
Enhanced EDGE
EFR
Enhanced Full Rate
FAM
FAn Module
FE
Fast Ethernet
FP
Frame Protocol
FR
Full Rate
FS
Fabric Switch
GE
Gigabit Ethernet
GERAN
GSM Edge Radio Access Network
GPS
Global Positioning System
GSM
Global System for Mobile communications
HR
Half Rate
HSPA+
HSPA Evolution
LTE
Long Term Evolution
MicroTCA
Micro Telecommunications Computing Architecture
MIMO
Multi Input Multi Output
MS/UE
Mobile Station/User Equipment
MTBF
Mean Time Between Failures
MTTR
Mean Time To Recovery
NBAP
Node B Application Part
NIS
Network Interface of STM-1
OPEX
OPeration EXpenditure
53
PM
Power Module
PS
Packet Switch
PtM
Point to Multi-point
PtP
Point to Point
RF
Radio Frequency
RNC
Radio Network Controller
RRU
Remote Radio Unit
RTT
Round Trip Time
SA
Site Alarm
SCS
System Control Sub-system
SDH
Synchronous Digital Hierarchy
SDR
Software Defined Radio
SE
Site Alarm Extension
TA
Time Advance
TAM
Tower mounted Amplifier control Module
TMA
Tower Mounted Amplifier
TNS
Transport Network Sub-system
UBPG
Universal Baseband Processing board for GSM
UBPG
Universal Baseband Processing board for GSM
UES
Universal Ethernet Switch Board
UL
Up Link
USB
Universal Serial Bus
UTRAN
UMTS Terrestrial Radio Access Network
WCDMA
Wideband Code Division Multiple Access
WiMAX
Worldwide Interoperability for Microwave Access
Reference Document
None
54

ZTE UMTS UR14 NodeB HSDPA Packet Scheduling Feature Guide

Uploaded by

Copyright:

Available Formats

ZTE UMTS UR14 NodeB HSDPA Packet Scheduling Feature Guide

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

ZTE UMTS UR14 NodeB HSDPA Packet Scheduling Feature Guide

Uploaded by

Copyright:

Available Formats

NodeB HSDPA Packet

NodeB HSDPA Packet Scheduling

NodeB HSDPA Packet Scheduling

Compared with UR13, chapter 6 was

2015 ZTE Corporation. All rights reserved.

ZTE Confidential Proprietary

NodeB HSDPA Packet Scheduling

Feature Attribute ............................................................................................... 5

Technical Description ....................................................................................... 8

Related Counters and Alarms ........................................................................ 38

ZTE Confidential Proprietary

NodeB HSDPA Packet Scheduling

ZWF23-05-002 HSDPA Flow Control ................................................................. 47

Engineering Guide .......................................................................................... 47

ZTE Confidential Proprietary

NodeB HSDPA Packet Scheduling

ZTE Confidential Proprietary

NodeB HSDPA Packet Scheduling

Performs cell-based fast scheduling for multiple UEs.

ZTE Confidential Proprietary

NodeB HSDPA Packet Scheduling

Transmits HS-SCCH control information and HS-DSCH data information.

Performs downlink HS-DSCH flow control.

ZWF23-01-013 HSDPA Fast Scheduling

Select a suitable HS-PDSCH air-interface transmission format according to the UE

ZWF23-05-002 HSDPA Flow Control

ZTE Confidential Proprietary

NodeB HSDPA Packet Scheduling

License Control List

Correlation With Other Features

ZTE Confidential Proprietary

NodeB HSDPA Packet Scheduling

ZWF23-01-013 HSDPA Fast Scheduling

HSDPA Resource Allocation Policies

HSDPA Power Resource Allocation Scheme

ZTE Confidential Proprietary

NodeB HSDPA Packet Scheduling

HSDPA Code Resource Allocation Scheme

HSDPA Scheduling Algorithms

ZTE Confidential Proprietary

NodeB HSDPA Packet Scheduling

MAX C/I Algorithm

ZTE Confidential Proprietary

NodeB HSDPA Packet Scheduling

RelativePriority =WeightofSPI Rate WeightofDelay (1+HistoryFlux)

ZTE Confidential Proprietary

NodeB HSDPA Packet Scheduling

Rate=w2(CQI_1) TBSIZE(CQI_1)+ w2(CQI_2) TBSIZE(CQI_2)

represents the service discardtime configured by the RNC.

Table 3-1 WeightofDelay Calculation

Tdiscard HOL ( ms)

ZTE Confidential Proprietary

NodeB HSDPA Packet Scheduling

The historical flux of a UE is calculated every 2 ms in accordance with the following

ZTE Confidential Proprietary

NodeB HSDPA Packet Scheduling

HSDPA TFRC Selection Algorithm

HS-SCCH Code and Power Selection

ZTE Confidential Proprietary

NodeB HSDPA Packet Scheduling