ZTE UMTS UR14 NodeB HSDPA Packet Scheduling Feature Guide
ZTE UMTS UR14 NodeB HSDPA Packet Scheduling Feature Guide
ZTE UMTS UR14 NodeB HSDPA Packet Scheduling Feature Guide
Scheduling
Feature Guide
Date
2015/03/30
Author
DuJuan,
Xuyun
Reviewer
Wangrui
Notes
TABLE OF CONTENTS
1
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
4
4.1
4.2
Parameters....................................................................................................... 36
ZWF23-01-013 HSDPA Fast Scheduling ........................................................... 36
ZWF23-05-002 HSDPA Flow Control ................................................................. 38
5
5.1
5.1.1
5.1.2
5.2
5.2.1
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
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.
The Adaptive Modulation and Coding (AMC) provides the HS-PDSCH air interface
transmission format (number of channelization codes, modulation mode, and TB
size).
Demodulates messages such as ACK, NACK and CQI carried on the HS-DPCCH.
2.1
Feature Introduction
2.1.1
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).
2.1.2
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
Feature ID
Feature Name
License
Configured
Control Item
NE
Unit
WCDMA
ZWF23-01-A
HSDPA
Package
Introduction
HSDPA
Introduction
RNC
CELL
Package
2.3
Technical Description
3.1
3.1.1
3.1.1.1
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
3.1.2
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
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:
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=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
WeightofDelay
< 10
65
[10, 20)
50
[20, 30)
37
[30, 40)
26
[40, 50)
17
[50, 60)
10
12
[60, 70)
[70, 80)
80
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
3.1.3.1
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:
3.1.3.2
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
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
is -2dB
Figure 3-1
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)
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
3.2
3.2.1
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
The number of stacked packets on the UE side is larger than the upper limit or is
smaller than the lower limit.
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
Trigger Event
Caculate 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
V in
represents the inflow rate over the lub interface of every UE, corresponding
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 ..
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
<
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
. If
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.
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
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
Transmitted carrier power of all codes not used for HS-PDSCH or HS-SCCH
transmission
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:
GbrTbSize = GBR/500, check the TS25.214 CQI table and get GbrCqi
according to GbrTbSize.
scheduled
times in 100 ms, check the TS25.214 CQI table, get SchCqi according to
SchTbSize.
= 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.
power
gdfMeasFilterCoeff,
thereinto
the
filtering
factor
gdfMeasFilterCoeff = 0.015625.
26
3.3
3.3.1
and
flow
control
algorithms.
But
it
has
certain
impact
on
refers
to
the
number
of
HS-PDSCH
channelization codes.
xccs,4, xccs,5, xccs,6, xccs,7 = |O-1-P/8 *15|
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
P8
13
P<8
24
8 10
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
29
3.3.2
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.
VAM technology.
2.
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:
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.
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
3.3.4
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:
3.3.5
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
3.3.7
the
technology
which
combines
DC
with
MIMO+64QAM
is
called
3.3.8
3.3.9
35
On carrier m, single-carrier UEs and MC UEs are queued and scheduled according to
the relative priorities.
3.3.10
3.3.11
Parameters
4.1
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
R,1:MAX None
C/I,2:PF
long:[1~6
]
None
36
Table 4-2
Managed
Object
Logic Name
Value
Parameter Description
Range
Unit
Default mende
Value
d
Value
UHspa
UMPO
UMPO
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
UMPO
Measurement
[1,2]
Configuration
of
UMPO
Configuration
of
of
UE
in
the
[0,1]
Cell(DC-HSDPA)
UHspa
RlcSizeSupt
Type
Flexible
PDU
Size [1 ,2,3]
Format
Downlink
uSize
[1~1505]; byte
Extended(bytes)
URncFunct CMaxPduSi
ion
UCHspa
ze
HSNBASSI
ND
Downlink
Conversational
Support [0,1]
none
Indicator
37
4.2
5.1
Related Counters
5.1.1
Counter ID
Name
C372490000
C372490001
C372490002
C372490003
C372490004
C372490005
C372490006
C372490007
C372490008
C372490009
C372490010
C372490011
C372490012
C372490013
C372490014
C372490015
C372490016
C372490017
38
C372490018
C372490019
C372490020
HSDPA Radio
C372490021
C372490022
C372490023
C372490024
C372490025
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
C372490073
C372490074
C372490075
C372490076
C372490077
C372490078
C372490079
C372490080
C372490081
C372490082
C372490083
C372490084
C372490085
C372490086
C372490087
C372490088
C372490089
C372490090
C372490091
C372490092
C372490093
C372490094
C372490095
C372490096
C372490097
40
C372490098
C372490099
C372490100
C372490101
C372490102
C372490103
C372490104
C372490105
C372490122
C372490123
C372490124
C372490125
C372490126
C372490127
C372490128
HSDPA Duration
C372490129
C372490130
C372490131
C372490132
C372490133
C372490134
C372490135
C372490136
C372490137
C372490138
C372490139
C372490140
C372490141
C372490142
41
C372490143
C372490144
C372490145
C372490146
C372490147
C372490148
C372490149
C372490150
C372490151
C372490152
C372490153
C372490160
C372490161
C372490162
C372490163
C372490164
C372490165
C372490166
C372490167
C372490168
C372490169
C372490170
C372490171
C372490172
42
C372490173
C372490174
C372490175
C372490176
C372490177
C372490178
C372490179
C372490180
C372490190
C372490191
C372490192
C372490193
C372490194
C372490198
C372490199
C372490200
C372490201
C372490202
C372490203
Throughput of MAC-EHS
C372490204
Scheduling of MAC-EHS
C372490205
C372490206
C372490207
C372490208
C372490209
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
C372490218
Throughput Rate of DC
C372490219
C372490220
C372490221
C372490222
C372490223
C372490224
C372490225
C372490226
C372490227
C372490228
C372490229
C372490230
C372490231
C372490232
C372490233
C372490234
C372490235
C372490236
C372490237
C372490238
C372490239
C372490240
C372490247
C372490248
C372490249
C372490250
44
C372490251
C372490252
C372490253
C372490254
C372490255
C372490256
C372490265
C372490266
C372490267
C372490268
C372490269
C372490270
C372490271
C372490272
C372490273
C372490274
C372490275
C372490276
C372490277
C372490278
C372490000
C372490001
C372490002
C372490315
45
C372490316
C372490317
C372490318
C372490319
C372490320
C372490321
C372490322
C372490323
C372490324
C372490325
C372490326
C372490327
C372490328
C372490329
C372490330
C372490331
C372490332
C372490333
C372490334
C372490335
46
5.1.2
5.2
Related Alarms
5.2.1
5.2.2
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
6.2.1
47
Figure 6-1
Figure 6-2
48
6.2.2
Figure 6-3
49
Figure 6-4
6.3
6.3.1
Test Item
1.
2.
Preconditio
3.
ns
4.
5.
Test Steps
Anticipative
Result
1.
2.
3.
1.
2.
50
6.3.2
2.
Preconditio
3.
ns
4.
5.
1.
2.
Test Steps
4.
1.
2.
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
6.4.1
GUI Name
Default
Deactivation
Value
Value
HSDPA
UUtranCellFDD.hsdStat
Function
1: Active
0: Inactive
Status
51
6.4.2
HSDPA Function
UUtranCellFDD.hsdStat
Default Value
GUI Name
Status
Deactivation
Value
1: Active
6.5
6.5.1
0: Inactive
None
6.5.2
Abbreviation
Abbreviation
Full Name
rd
3GPP
64QAM
AGPS
Assisted GPS
AMR
BB
Base Band
BBU
BCCH
BITS
BPK
BSC
BSP
BTS
CAPEX
CAPital EXpenditure
52
CC
CE
Channel Element
CN
Core Network
CORBA
CPRI
CS
Circuit Switch
CS
Carrier Sector
DL
Down Link
DTX
Discontinuous transmission
EDGE
E-EDGE
Enhanced EDGE
EFR
FAM
FAn Module
FE
Fast Ethernet
FP
Frame Protocol
FR
Full Rate
FS
Fabric Switch
GE
Gigabit Ethernet
GERAN
GPS
GSM
HR
Half Rate
HSPA+
HSPA Evolution
LTE
MicroTCA
MIMO
MS/UE
MTBF
MTTR
NBAP
NIS
OPEX
OPeration EXpenditure
53
PM
Power Module
PS
Packet Switch
PtM
Point to Multi-point
PtP
Point to Point
RF
Radio Frequency
RNC
RRU
RTT
SA
Site Alarm
SCS
SDH
SDR
SE
TA
Time Advance
TAM
TMA
TNS
UBPG
UBPG
UES
UL
Up Link
USB
UTRAN
WCDMA
WiMAX
Reference Document
None
54