July 2007
doc.: IEEE 802.11-07/2187r1
Another resource to exploit: multi-user diversity
Authors:
Date: 2007-07-17
Name
Company Address
Phone
email
Marc de Courville
Motorola
+33169352518
[email protected]
m
[email protected]
m
[email protected]
+33169352564
[email protected]
Mohamed Kamoun
Anahid Robert
Jeremy Gosteau
Parc les
Algorithmes
Saint Aubin,
91193 Gif
sur Yvette
Cedex
Roberta Fracchia
[email protected]
Sophie Gault
[email protected]
Notice: This document has been prepared to assist IEEE 802.11. It is offered as a basis for discussion and is not binding on the contributing individual(s) or organization(s). The material in
this document is subject to change in form and content after further study. The contributor(s) 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 802.11.
Patent Policy and Procedures: The contributor is familiar with the IEEE 802 Patent Policy and Procedures <http:// ieee802.org/guides/bylaws/sb-bylaws.pdf>, including the statement
"IEEE standards may include the known use of patent(s), including patent applications, provided the IEEE receives assurance from the patent holder or applicant with respect to patents
essential for compliance with both mandatory and optional portions of the standard." Early disclosure to the Working Group of patent information that might be relevant to the standard is
essential to reduce the possibility for delays in the development process and increase the likelihood that the draft publication will be approved for publication. Please notify the Chair
[email protected] as early as possible, in written or electronic form, if patented technology (or technology under patent application) might be incorporated into a draft standard being
developed within the IEEE 802.11 Working Group. If you have questions, contact the IEEE Patent Committee Administrator at <[email protected]>.
Submission
Slide 1
Marc de Courville (Motorola)
July 2007
doc.: IEEE 802.11-07/2187r1
Motivation for 11vht multi band evolution
• Reality of the market:
– Wi-Fi penetration in the home/enterprise/outdoor is growing
– Need to face diversity of device classes and features in a robust way
– Need to support multiple connections: throughput shouldn’t drop when increasing the
number of devices (in the home simultaneous support for IPTV/MP3 streaming & VoIP
calls is a reality)
– All the more true for public access networks and office/enterprise communication
solutions
• Core questions: How to support various BW devices without too much
overhead? Is there a value to have several devices transmitting at the same
time?
– OFDMA exists and provides solutions to:
• Unbalanced link budgets
• Improved spectral efficiency/higher capacity through more optimal resource scheduling
– Dynamic adaptive bandwidth enables spectrum allocation to be sized to traffic need
Submission
Slide 2
Marc de Courville (Motorola)
July 2007
doc.: IEEE 802.11-07/2187r1
Padding loss: it increases with the bandwidth!
• Fact: when increasing the bandwidth the OFDM
const
granularity becomes quite large and even larger with
BPSK
MIMO to accommodate various packet sizes resulting in QPSK
QPSK
increased padding inefficiencies!
16-QAM
• Average: bit per carrier considering coverage of MCSs 16-QAM
64-QAM
(SISO TGnD, 5% PER): around 1.47bit/carrier
64-QAM
• Padding loss on the useful message versus bandwidth
64-QAM
utilization: 1.7dB (average) and up to 2.2dB for VoIP
packet (ACELP AMR codec at 12.2kbps, 20ms sampling;
counting RTP+UDP+IP=40B header and 11 MAC
header=34B)
BW
20
40
80
100
size (B)
FFT side P
64
4 4
128
7 6
256 10 8
320 12 10
105 (VoIP)
Z
1
3
5
7
U
52
106
224
280
size (B)
min
max
3,3 32,5
6,6 66,3
14,0 140,0
17,5 175,0
av
9,5
19,4
41,1
51,3
#OFDM
max
min av
32,2 3,2 11,0
15,8 1,6 5,4
7,5 0,7 2,5
6,0 0,6 2,0
R
1/2
1/2
3/4
1/2
3/4
2/3
3/4
5/6
coverage bits
26,0%
0,5
38,7%
1
0,1%
1,5
18,2%
1,5
4,1% 2,25
4,5%
4
3,4%
4,5
4,9%
5
1,47
Padding loss(% and dB)
max min
av
max min
2,6% 19,6% 0,3% -0,1 -0,9
1,4% 21,1% 10,3% -0,1 -1,0
6,7% 25,4% 15,1% -0,3 -1,3
0,5% 40,3% 32,1% 0,0 -2,2
av
0,0
-0,5
-0,7
-1,7
• Conclusion: with increased granularity we can gain some dBs…
Submission
Slide 3
Marc de Courville (Motorola)
July 2007
doc.: IEEE 802.11-07/2187r1
Materializing multi-user diversity gain
• In order to increase granularity let assume several STAs are allowed to
transmit at the same time with an orthogonal divisions of subcarriers
(OFDMA).
• Goal: determine upper bound provided by frequency selective scheduling a.k.a.
multi-user diversity gain
• Assumptions: OFDMA, frequency multiplexing of the users
– SISO TGnD channel TRMS=50ns, coherent bandwidth Bc=20MHz
thus as a rule of thumb channel is almost constant over 12
consecutive subcarriers (Bc/5).
– Same carrier spacing as for 20MHz is assumed: 312.5kHz
– 100MHz bandwidth representing 320 subcarriers
– 8 adjacent subcarriers are grouped into one chunk
– Nu users are present: 320/Nu/8 chunks are allocated to each user
– Chunks are chosen so as to maximize the sum rate capacity
• Acronyms:
– OCA: optimized chunk allocation
– RCA: random chunk allocation
Submission
Slide 4
Marc de Courville (Motorola)
July 2007
doc.: IEEE 802.11-07/2187r1
PHY over the air throughput
6 users: 66% rate increase
2 users: 33% rate increase
Submission
Slide 5
Marc de Courville (Motorola)
July 2007
doc.: IEEE 802.11-07/2187r1
Equivalent SNR Gain of OCA vs. RCA function of SNR
>2dB gain already with 2 users!
Submission
Slide 6
Marc de Courville (Motorola)
July 2007
doc.: IEEE 802.11-07/2187r1
PHY Rate CDF
~50Mbps
>75Mbps
~50Mbps
Submission
Slide 7
Marc de Courville (Motorola)
July 2007
doc.: IEEE 802.11-07/2187r1
Some conclusions on the simulations and implications
• Simulation results:
– BTW OCA represents also a one user full bandwidth usage
– Multi-User diversity gain starts with a 2dB for 2 users served and reaches 5dB
for 6 users
– In terms of throughput benefit: at 10dB SNR, 33% gain with 2 users and 66%
gain with 6 users
Submission
Slide 8
Marc de Courville (Motorola)
July 2007
doc.: IEEE 802.11-07/2187r1
Next steps and challenges
•
Thoughts on link budget: is the OFDMA power boost required?
– AP 1W limitation (30dBm: 5GHz upper band), STA 200mW (SAR regulation 23dBm)
– DL: 100MHz (5x20MHz) yields same power spectral density as for 200mW 20MHz transmissions
– UL: 200mW limitation for STA indicates an unbalanced link budget that can be compensated with
OFDMA serving 5 users increasing the energy per subcarrier by a factor of 5
•
•
Challenge: this presentation deals only with a potential PHY gain to be materialized by
dedicated MAC protocols
Open questions:
– What are the required ingredients to feed the PAR that will allow to capitalize on this new source of
diversity?
– Can OFDMA be coupled with smart interference management to handle overlapping BSS for
efficient deployments?
– From SISO to MIMO: does the frequency selective scheduling gain remain due to the “implicit
antenna averaging”? It might depend on the MIMO mode.
•
Conclusion:
– Multi-user diversity is a resource not exploited by IEEE802.11n
– OFDMA is a proven technology already present in IEEE802.16, 3GPP LTE is now investigating
Multi-User MIMO (e.g. PU2RC).
– IEEE802.11vht should take these existing and new trends into consideration in the PAR and in the
specification drafting process.
Submission
Slide 9
Marc de Courville (Motorola)
July 2007
doc.: IEEE 802.11-07/2187r1
PAR recommendations
• Adopt a requirement on the aggregated throughput and not the
peak throughput to introduce multi-user component into the
standard
Submission
Slide 10
Marc de Courville (Motorola)
July 2007
doc.: IEEE 802.11-07/2187r1
Backup: influence of the chunk size
• Context: comparison with chunks
of size 4, 8, 32
• Slight improvement with smaller
chunks but 10 chunks of 32
carriers would require less MAC
overhead
Submission
Slide 11
Marc de Courville (Motorola)