E-MBS Subburst Partition (16.3.6.5.3 and 16.9.1)
Document Number: IEEE S802.16m-10/0071
Date Submitted: 2010-3-4
Source:
Zheng Yan-Xiu, Yu-Chuan Fang, Chun-Yuan Chiu, Chang-Lan Tsai, Chung-Lien Ho
ITRI
E-mail: zhengyanxiu@itri.org.tw
Venue:
IEEE 802.16 Session#66 at Orlando, USA.
Base Contribution:
C802.16m-10/0071
Purpose:
Propose to be discussed and adopted by TGm for the use in Project 802.16m/D4
Notice:
This document does not represent the agreed views of the IEEE 802.16 Working Group or any of its subgroups. It represents only the views of the participants listed in
the “Source(s)” field above. It is offered as a basis for discussion. It is not binding on the contributor(s), who reserve(s) the right to add, amend or withdraw material
contained herein.
Release:
The contributor grants a free, irrevocable license to the IEEE to incorporate material contained in this contribution, and any modifications thereof, in the creation of an
IEEE Standards publication; to copyright in the IEEE’s name any IEEE Standards publication even though it may include portions of this contribution; and at the IEEE’s
sole discretion to permit others to reproduce in whole or in part the resulting IEEE Standards publication. The contributor also acknowledges and accepts that this
contribution may be made public by IEEE 802.16.
Patent Policy:
The contributor is familiar with the IEEE-SA Patent Policy and Procedures:
<http://standards.ieee.org/guides/bylaws/sect6-7.html#6> and <http://standards.ieee.org/guides/opman/sect6.html#6.3>.
Further information is located at <http://standards.ieee.org/board/pat/pat-material.html> and <http://standards.ieee.org/board/pat >.
E-MBS Resource Allocation
•
•
•
•
In general, an E-MBS burst is very large and spans on multiple frames or
superframes.
Large E-MBS burst implies multiple FEC blocks necessary to carry these information
bits.
Based on current D4, either one of these FEC blocks is incorrect, a complete E-MBS
burst is discarded.
– The larger number of FEC blocks, the higher error rate
In order to avoid high error rate, we shall partition a burst into multiple subbursts
–
•
•
If several FEC blocks are incorrect, other correct FEC blocks still can passed to higher layer
Furthermore, the maximum burst, 14400 bytes, may not contain complete E-MBS
burst and subburst partition is still necessary
If higher layer coding is utilized, the coding performance does not suffers from
complete E-MBS burst error and smaller FEC blocks favors upper layer coding
…
CRC encoder
CRC encoder
…
FEC encoder
…
FEC encoded blocks
FEC encoder
…
FEC encoded blocks
mapped into
resource units
LRUs
…
mapped into
resource units
LRUs
LRUs
LRUs
…
Higher Layer Coding Support
•
Upper layer coding is used for EMBS service to enhance
performance
–
•
•
•
•
•
Fountain coding, Raptor codes,
LDPC cods, RS codes, etc.
The upper layer views packets
from lower layer through error and
erasure channel
In general, extra 10% redundancy
is enough decode data
With the higher layer coding
support, the error rate can be
highly reduced
However, larger FEC block size
implies smaller code length for
upper layer coding and decreases
error rate performance
In order to support higher layer
coding, smaller FEC block size
might be considered
data blocks
A
B
C
…
Upper layer
coding
…
RU-based blocks
X
Y
Z
…
other functions
…
…
…
…
…
CRC encoder
…
FEC encoder
…
FEC encoded blocks
mapped into
resource units
LRUs
LRUs
LRUs
…
one scheduling interval
3
3
Simulation Parameter
Parameters
Assumption
Bandwidth
10 MHz
Number of subcarrier
1024
Frame length
5ms
Number of DL subframe per frame
5
Number of resource unit per subframe
20
Channel estimation
2D-MMSE
Channel code
CTC 1/2
Modulation
QPSK
FEC block size (Nep)
240, 480, 960, 1920, 3840 (bits)
MIMO configuration
SM, Tx: 2, Rx:2
Channel model
Ped-B 3km/h, Veh-A 30km/h
BLER: FEC block error rate
SFER: super-frame error rate
4
Part I: Burst vs. Subburst
• The error rate increases with velocity
– For PB 3km/hr case, 1dB gap is on 10-3
– For VA 30km/hr case, 2dB gap is on 10-3
– When mobility becomes higher, time domain correlation
becomes less
– If larger more superframes, e.g. 8 superframes = 160ms,
are considered, the higher error rate probability
• Burst partition shall be considered
• In order to reduce complexity, the partitioned subburst
shall be aligned with the size of multiple LRUs.
6
7
Part II: Subburst with Higher Layer Coding
• Analysis performance is provided
– Upper layer coding
• Consider there E-MBS data burst is partitioned into N packets
• The upper layer coder encode the N packets into M packets
– Assumption: if extra 10% redundant packets are received on
higher layer, the entire E-MBS burst is correct
M 
M
k
M k
P

– We apply the equation e,EMBS Burst k 1.1N 1  k  Pe,FEC (1  Pe,FEC )
 
to analyze the error rate performance with higher layer code,
where Pe,E-MBS Burst is the error rate of E-MBS burst, Pe,FEC is the
error rate of FEC burst
• Results: the larger the FEC block size, the smaller number of N
– When 30% overhead is considered, 240 bits FEC block performs
the best
• Recommendation: Subband-based partition may be consdiered for
E-MBS burst partition
9
10
11
12
Part III: E-MBS with Upper Layer Coding
• E-MBS burst error rate comparison is shown
• With upper layer coding, up to 7dB@Error Rate=10-3
performance can be seen with 100 % redundancy
– If we consider deeper error rate, more performance gain
can be achieved due to sharper slope of error rate curve, it
might be up to 14 dB @ Error Rate = 10-6
• The upper layer coding improve end-to-end performance
and the associated QoE
• 16m should support higher layer coding through
subburst partition
14
15
16
17
Conclusions
• E-MBS burst shall be partitioned into subbursts to
reduce error rate.
• E-MBS burst can be partitioned into subbursts given the
length fitting to subband size
– No more than the biggest FEC blocks size
– The minimum number of bits is around 392 for QPSK code
rate ½
– It can follow the existing FEC coding procedure
• With the subburst partition support, the performance can
be improved
• The design is transparent to any upper layer coding
• Proposed text see in contribution IEEE C802.16m10/0071 or later versions.
References
• [1] 3GPP TS 25.346 V8.1.0 (2008-03), “Introduction of the
Multimedia Broadcast Multicast Service (MBMS) in the Radio
Access Network (RAN); Stage 2 (Release 8),” Mar. 2008.
• [2] 3GPP TS 26.346 V7.8.0 (2008-06), “Multimedia
Broadcast/Multicast Service (MBMS); Protocols and codecs
(Release 7),” Jun. 2008.
• [3] IEEE P802.16m/D3, “Draft Amendment to IEEE Standard for
Local and Metropolitan Area Networks—Part 16: Air Interface for
Broadband Wireless Access,” Dec. 2009.
19
19
Appendix A: Performance of E-MBS
Burst Error Rate w/ and w/o Upper
Layer Coding
21
22
23
24
25
26
27
28
29
30
31
32
Appendix B: Performance of E-MBS
Burst Error Rate w/ and w/o Upper
Layer Coding
34
35
36
37
38
39
40
41
42
43
44
45