HARQ Feedback AMAP Design
Document Number: IEEE C80216m-09/1029
Date Submitted: 2009-4-27
Source:
Debdeep Chatterjee, Yi Hsuan, Hujun Yin Email: {[email protected], [email protected], [email protected]}
Intel Corporation
Hyunkyu Yu, Jaeweon Cho, Heewon Kang, Hokyu Choi Email: {[email protected] }
Samsung
Hyungho Park, Jinsoo Choi, Bin-chul Ihm, Jin Sam Kwak Email: {hyunghopark, emptylie, bcihm, samji}@lge.com
LGE
Venue:
IEEE Session #61, Cairo, Egypt.
Re: AWD comments / Area: Chapter 15.3.6 (DL-CTRL), “Comments on AWD 15.3.6 DL-CTRL”
Base Contribution:
N/A
Purpose:
For TGm discussion and adoption of 802.16m AWD text.
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 >.
1
HARQ Feedback AMAP: Performance Requirements
• Maximum error rates for HARQ Feedback AMAP:
– Pr [NAK  ACK] = 0.1%
– Pr [ACK  NAK] = 1%
• For 95% coverage for baseline cell size of 1.5km ISD:
– Transmission with error rates not exceeding the above at an SINR as
low as -3.7dB
• For 95% coverage for larger cell size of 5km cell-radius:
– Transmission with error rates not exceeding the above at an SINR as
low as -6dB
2
HARQ Feedback AMAP: Proposed Design
HF-A-MAP IE(s)
[2 bits if QPSK;1 bit if BPSK]
Repetition
(x8)
QPSK/
BPSK
SFBC
HF-A-MAP IE(s)
[2 bits if QPSK;1 bit if BPSK]
Repetition
(x8)
QPSK/
BPSK
HF-AMAP
symbols
• Each HF-A-MAP IE carries 1 bit information
• Depending on the channel conditions, the modulation can be QPSK
or BPSK
– If QPSK is used, then 2 ACK IEs are mapped to a point in the signal
constellation
– If BPSK is used, then each ACK IE is mapped to a point in the signal
constellation
• Frequency resource usage: 4 or 8 tones per user (i.e. HF-A-MAP IE)
corresponding to QPSK or BPSK modulation
3
Simulation Parameters
Total available bandwidth
10 MHz (1024 subcarriers)
Carrier Frequency
2.5GHz
Number of OFDM symbols per subframe
6
Number of total RU in one subframe
48
Number of Antennas
2 transmitter antennas, 2 receiver antennas [2Tx,2Rx]
MIMO mode
2x2 SFBC
Tone selection
Fully distributed (uniformly) over the entire band in units of tone-pairs (SFBC)
Modulation/Coding
QPSK or BPSK, with 8 repetitions in the frequency domain
MMSE with White Noise Assumption (MMSE-WNA) with perfect SINR at
MIMO Receiver
Rx
MMSE with Colored Noise Consideration (MMSE-CNC)
Traffic model
eITU-VehA 120 km/h
Pilot patterns
2-stream pilot pattern as in SDD
Pilot boosting
Each Tx. antenna boosts its pilot tone by 5dB
Channel estimation
2-D MMSE PRU-based channel estimator
Scenarios
Noise-limited (NL): SIR = Inf
Interference-limited (IL): INR = 10dB, 2 interferers, interfering data
symbols are always QPSK
Probability of error (BER) operating points
0.1% probability for both ACK  NACK and NACK  ACK errors.
4
The MMSE-CNC Receiver (1/2)
• MMSE receiver that estimates the noise + interference power from
the pilot tones and performs MMSE detection/equalization with
colored noise consideration
• Unlike the MMSE-WNA Rx, perfect knowledge of SINR at the
receiver is not assumed. The receiver estimates the noise power
from the pilots
– Causes some degradation in the noise-limited performance, but
this is more realistic
• The MMSE equalization is performed using the estimated noise
covariance matrix (that may be colored in the presence of
interference), instead of the white noise assumption (for which the
noise covariance matrix is diagonal)
– Improves the performance in interference-limited scenarios
5
The MMSE-CNC Receiver (2/2)
•
Received symbol:
•
Noise covariance estimated from the pilots averaged over a particular PRU
(or over 3 PRUs when wideband channel estimation is supported):
•
For a particular data tone
is
•
Proceed with the standard MMSE detection/equalization with
covariance matrix for data tone
, the estimated noise covariance matrix
as the noise
6
Noise-Limited (NL): QPSK Modulation, MMSE-WNA Rx
(Perfect Noise Covariance from SNR at Rx)
7
NL: QPSK Modulation, MMSE-CNC Rx
8
NL: BPSK Modulation, MMSE-WNA Rx (Perfect Noise
Covariance from SNR at Rx)
9
NL: BPSK Modulation, MMSE-CNC Rx
10
Interference-Limited (IL): QPSK Modulation,
MMSE-WNA Rx, Interfering data = QPSK symbols
11
IL: QPSK Modulation, MMSE-CNC Rx, Interfering data =
QPSK symbols
12
IL: BPSK Modulation, MMSE-WNA Rx, Interfering data =
QPSK symbols
13
IL: BPSK Modulation , MMSE-CNC Rx, Interfering data
= QPSK symbols
14
Summary and Conclusion
• The MMSE-CNC Rx results in approximately 0.75~1dB degradation (at 0.1%
BER) when compared to the MMSE-WNA Rx with perfect SNR at the receiver for
NL scenarios
• The MMSE-CNC Rx provides gains of about 2.4~2.8dB (at 0.1% BER) over the
MMSE-WNA Rx for IL scenarios
• The data power boosting level can be limited to 4~5dB by appropriately
choosing the MCS (QPSK or BPSK) according to user channel conditions.
–
ACK IEs for users with very poor channel conditions may be transmitted using
BPSK modulation instead of increased data power boosting
• Appropriate MCS adaptation (as above) potentially avoid high amounts of data
power boosting or de-boosting
–
–
Very high data power boosting may cause high co-channel interference
Severe data power de-boosting, on the other hand, may cause the de-boosted
data tones to be very sensitive to effects of inter-channel interference (ICI)
• Even further performance gains may be realized by exploiting the fact that the
ACK  NAK and NAK  ACK error probability requirements are unequal
–
A suitably chosen non-zero threshold may be used at the threshold detection
block in the receiver (the hard decision step) at the receiver
15
Proposed AWD Text Change:
Replace Figure 428 on page 53 with the following figure
HF-A-MAP IE(s)
[2 bits if QPSK;1 bit if BPSK]
Repetition
QPSK/
BPSK
SFBC
HF-A-MAP IE(s)
[2 bits if QPSK;1 bit if BPSK]
Repetition
QPSK/
BPSK
HF-A-MAP
symbols
and add the following text after line 58.
“Each HF-A-MAP IE carries 1 bit information. Depending on the channel
conditions, the modulation can be QPSK or BPSK. If QPSK is used, 2
HF-A-MAP IEs are mapped to a point in the signal constellation. If BPSK
is used, each HF-A-MAP IE is mapped to a point in the signal
constellation. The repetition number is FFS.”
16