DL Partitioning and Permutation Evaluation
Document Number: IEEE C802.16m-09/0251
Date Submitted: 2009-01-07
Source: Furuzan Atay, Rongzhen Yang, Jong-Kae (JK) Fwu, Minh-Anh Vuong, Yuval Lomnitz, Huaning Niu, Hujun Yin
Email: jong-kae.fwu@intel.com
Venue: IEEE Session #59, San Diego.
Base Contributions:
None
Re: IEEE C802.16m-08/053 “Call for Comments and Contributions on Project 802.16m Amendment Working Document ” , Section 15.3.5
Downlink Physical Structure
Purpose: Discussion and Approval
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 >.
Motivation
• Different proposal for subband partitioning, miniband
permutation and subcarrier permutation from different
companies in IEEE C802.16m-08/1508r1
• Performance evaluation and comparison are needed to
make a decision
Evaluation Scenarios
BW = 10 MHz, NPRU = 48 (N1 = 4, N2 = 1)
Scenario1 (Mixed Reuse 1&3 – equal size)
Scenario 3 (Reuse 1)
# of subbands
(KSB,FP,i)
# of minibands
(KMB,FPi)
# of PRUs in
FPi
FP1
2
4
12
FP2
2
4
FP3
2
FP4
2
Freq.
Partition
# of subbands
(KSB,FP,i)
# of
minibands
(KMB,FPi)
# of PRUs in
FPi
FP1
6
24
48
12
FP2
0
0
0
4
12
FP3
0
0
0
4
12
FP4
0
0
0
# of subbands
(KSB,FP,i)
# of
minibands
(KMB,FPi)
# of PRUs in
FPi
Scenario 2 (Mixed Reuse 1&3 – unequal size)
Freq.
Partition
Scenario 4 (Reuse 3)
# of subbands
(KSB,FP,i)
# of minibands
(KMB,FPi)
# of PRUs in
FPi
FP1
4
8
24
FP1
0
0
0
FP2
0
8
8
FP2
2
8
16
FP3
0
8
8
FP3
2
8
16
FP4
0
8
8
FP4
2
8
16
Freq.
Partition
Freq.
Partition
Evaluation Criteria Description
• Diversity gain for distributed resources
– Calculate the equivalent SNR of each LRU and plot the
histogram
– Find 10% outage SNR, i.e, the SNR value which can be
achieved with outage 0.1.
• Scheduling gain for contiguous resources
– Calculate the equivalent SNR of each LRU and select the best
LRU for each FFR. Plot the histogram of the best SNR.
– Find 10% outage SNR, i.e, the (best) SNR value which can be
achieved with outage 0.1.
• Hit ratio for distributed resources
– Average number of subcarrier pairs hit by neighboring cell.
DL Permutation Demonstration Model
kth DLRU at BS
17
33
65
81
18
34
66
82
1
19
35
49
67
83
2
20
36
50
68
3
21
37
51
4
22
38
5
23
6
24
7
25
8
26
9
27
41
10
28
11
kth DLRU at MS
17
33
65
81
18
34
66
82
1
19
35
49
67
83
84
2
20
36
50
68
84
69
85
3
21
37
51
69
85
52
70
86
4
22
38
52
70
86
39
53
71
87
5
23
39
53
71
87
40
54
72
88
6
24
40
54
72
88
55
73
7
25
55
73
56
74
8
26
56
74
57
75
89
9
27
41
57
75
89
42
58
76
90
10
28
42
58
76
90
29
43
59
77
91
11
29
43
59
77
91
12
30
44
60
78
92
12
30
44
60
78
92
13
31
45
61
79
93
13
31
45
61
79
93
14
32
46
62
80
94
14
32
46
62
80
94
15
47
63
95
15
47
63
95
16
48
64
96
16
48
64
96
NRx=2
MMSE
Receiver
E-ITU channel
PedB, 3km/h
54
Store sum
rate of all
streams
Rate of each stream is
computed via Shannon
formula by using
post-SNR
NTx=2
PTx(dBm)=10+Pnoise(dBm)
PTxPerAntenna(mW)=
PTx(mW)/NTx
SNR1 of DLRU1
Plot CDF of all SNRi
1
0.9
0.8
0.7
CDF
0.6
0.5
SNRk of DLRUk
0.4
DLRU: Distributed LRU
0.3
0.2
0.1
6
8
10
0.1
12
14
16
SNR(dB)
18
20
*** 10% outage SNR ***
22
24
26
SNRN of DLRUN
Compute average
rate of
all tones
Compute the equivalent SNR
from this average rate by
reversing Shannon formula
Part I
Comparison of Subband Partitioning and
Miniband Permutation Proposals
Comparison of Subband Partitioning
and Miniband Permutations
• Compare Intel1, Intel2, LGE, Samsung, Motorola
• Same subcarrier perm used for all proposals
– perm_seq() is random permutation
– Shift by symbol and subframe index according to
Intel’s formula.
Study 1: 10% SNR Values
Scenario1 (Mixed Reuse 1&3 – equal size)
Proposals
10% SNR dB
Distributed
10% SNR dB
Contiguous
Intel1
11.1831
15.3912
Intel2
11.2824
LGE
Scenario 3 (Reuse 1)
Proposals
10% SNR dB
Distributed
10% SNR dB
Contiguous
Intel1
12.1394
18.9974
15.2291
Intel2
12.0804
18.8838
11.1738
15.3395
LGE
12.1186
18.9178
Samsung
11.1812
15.3928
Samsung
12.0671
19.0390
Motorola
11.2818
14.9181
Motorola
12.1340
18.7412
Scenario 2 (Mixed Reuse 1&3 – unequal size)
Proposals
10% SNR dB
Distributed
10% SNR dB
Contiguous
Scenario 4 (Reuse 3)
Proposals
10% SNR dB
Distributed
10% SNR dB
Contiguous
Intel1
11.8191
17.9200
Intel1
11.7858
11.8128
Intel2
11.8034
18.0018
Intel2
11.7797
11.8082
LGE
11.9008
17.9618
LGE
11.8676
11.9212
Samsung
11.8293
17.9686
Samsung
11.7922
11.8128
Motorola
11.5811
17.6156
Motorola
11.6648
11.7076
Diversity and scheduling gain in
Scenario 1
Scenario 1, Distributed resources, 2x2 MIMO
Scenario 1, Contiguous resources, 2x2 MIMO
1
1
SB Part=Intel1, MB Perm=Intel
SB Part=Intel2, MB Perm=Intel
SB Part=LGE, MB Perm=LGE
SB Part=Samsung, MB Perm=Samsung
SB Part=Motorola, MB Perm=Motorola
0.9
0.8
0.7
0.8
0.7
0.6
CDF
CDF
0.6
0.5
0.5
0.4
0.4
0.3
0.3
0.2
0.2
0.1
0.1
0
SB Part=Intel1, MB Perm=Intel
SB Part=Intel2, MB Perm=Intel
SB Part=LGE, MB Perm=LGE
SB Part=Samsung, MB Perm=Samsung
SB Part=Motorola, MB Perm=Motorola
0.9
5
10
15
SNR(dB)
20
25
0
5
10
15
20
SNR(dB)
25
30
35
Diversity and scheduling gain in
Scenario 2
Scenario 2, Contiguous resources, 2x2 MIMO
1
0.9
0.9
0.8
0.8
0.7
0.7
0.6
0.6
CDF
CDF
Scenario 2, Distributed resources, 2x2 MIMO
1
0.5
0.5
0.4
0.4
0.3
0.3
0.1
0
SB Part=Intel1, MB Perm=Intel
SB Part=Intel2, MB Perm=Intel
SB Part=LGE, MB Perm=LGE
SB Part=Samsung, MB Perm=Samsung
SB Part=Motorola, MB Perm=Motorola
SB Part=Intel1, MB Perm=Intel
0.2
SB Part=Intel2, MB Perm=Intel
SB Part=LGE, MB Perm=LGE
SB Part=Samsung, MB Perm=Samsung 0.1
SB Part=Motorola, MB Perm=Motorola
0.2
6
8
10
12
14
16
SNR(dB)
18
20
22
24
0
12
14
16
18
20
22
SNR(dB)
24
26
28
30
32
Diversity and scheduling gain in
Scenario 3
Scenario 3, Distributed resources, 2x2 MIMO
Scenario 3, Contiguous resources, 2x2 MIMO
1
1
SB Part=Intel1, MB Perm=Intel
SB Part=Intel2, MB Perm=Intel
SB Part=LGE, MB Perm=LGE
SB Part=Samsung, MB Perm=Samsung
SB Part=Motorola, MB Perm=Motorola
0.9
0.8
0.7
0.9
0.8
0.7
0.6
CDF
CDF
0.6
0.5
0.5
0.4
0.4
0.3
0.3
0.2
0.2
0.1
0.1
0
SB Part=Intel1, MB Perm=Intel
SB Part=Intel2, MB Perm=Intel
SB Part=LGE, MB Perm=LGE
SB Part=Samsung, MB Perm=Samsung
SB Part=Motorola, MB Perm=Motorola
8
10
12
14
16
SNR(dB)
18
20
22
24
0
10
15
20
25
SNR(dB)
30
35
Diversity and scheduling gain in
Scenario 4
Scenario 4, Distributed resources, 2x2 MIMO
Scenario 4, Contiguous resources, 2x2 MIMO
1
1
SB Part=Intel1, MB Perm=Intel
SB Part=Intel2, MB Perm=Intel
SB Part=LGE, MB Perm=LGE
SB Part=Samsung, MB Perm=Samsung
SB Part=Motorola, MB Perm=Motorola
0.9
0.8
0.7
0.9
0.8
0.7
0.6
CDF
CDF
0.6
0.5
0.4
0.4
0.3
SB Part=Intel1, MB Perm=Intel
SB Part=Intel2, MB Perm=Intel
SB Part=LGE, MB Perm=LGE
SB Part=Samsung, MB Perm=Samsung
SB Part=Motorola, MB Perm=Motorola
0.3
0.2
0.2
0.1
0
0.5
0.1
6
8
10
12
14
16
SNR(dB)
18
20
22
24
0
5
10
15
20
SNR(dB)
25
30
35
Conclusions of Part I
• Proposals from Intel (1&2), LGE, and
Samsung have the best performance in
terms of diversity and scheduling gain.
Part II
Comparison of Subcarrier Permutation
Proposals
Comparison of Subcarrier Permutation
Proposals
• Compare subcarrier permutation proposed by Intel, LGE,
Samsung, and random permutation
• Use Intel1 SB partitioning and Intel MB permutation
• Random permutation: PermSeq is random, shifted over
symbols and subframes according to Intel’s formula.
• Calculate diversity gain for different subcarrier perms for
Scenarios 1, 2, and 3.
Diversity gain: 10% SNR Values
Scenario1 (Mixed Reuse 1&3 – equal size)
Proposals
10% SNR dB
Distributed
Intel
11.2333
Samsung
11.2297
LGE
11.1530
Random
11.2129
Scenario 2 (Mixed Reuse 1&3 – unequal size)
Proposals
10% SNR dB
Distributed
Intel
11.8724
Samsung
11.7254
LGE
11.6638
Random
11.7934
Scenario 3 (Reuse 1)
Proposals
10% SNR dB
Distributed
Intel
12.1715
Samsung
12.1740
LGE
11.4519
Random
12.0221
Diversity gain in Scenario 1
Scenario 1, Distributed resources, 2x2 MIMO
1
0.9
0.8
0.7
CDF
0.6
0.5
0.4
SubCarPerm=Intel
SubCarPerm=Samsung
SubCarPerm=LGE
SubCarPerm=random
0.3
0.2
0.1
0
6
8
10
12
14
16
SNR(dB)
18
20
22
24
Diversity gain in Scenario 2
Scenario 2, Distributed resources, 2x2 MIMO
1
SubCarPerm=Intel
SubCarPerm=Samsung
SubCarPerm=LGE
SubCarPerm=random
0.9
0.8
0.7
CDF
0.6
0.5
0.4
0.3
0.2
0.1
0
6
8
10
12
14
SNR(dB)
16
18
20
22
Diversity gain in Scenario 3
Scenario 3, Distributed resources, 2x2 MIMO
1
0.9
0.8
0.7
CDF
0.6
SubCarPerm=Intel
SubCarPerm=Samsung
SubCarPerm=LGE
SubCarPerm=random
0.5
0.4
0.3
0.2
0.1
0
8
10
12
14
16
SNR(dB)
18
20
22
24
Conclusions of Part II
• Subcarrier permutation methods proposed
by Intel and Samsung have the best
diversity gain.
Part III
Comparison of Average Hit Number
DL Hit Number Comparison
• Compare the number of hits among different Cell ID
pairs for the subcarrier permutations proposed by Intel,
Samsung and LGE.
• 0≤ Cell_ID <512, 130816 Cell_ID Pairs
• Intel
– PermSeq remains the same for all cells.
• Samsung and LTE:
– Different PermSeq for different cells:
– LGE: different shift as function of Cell_ID
– Samsung: different SEED as function of Cell_ID
Hit Number Comparison
• k: number of subcarrier pairs hit within an LRU,
(0 ≤ k ≤ kmax and kmax = Nsym*Lpair = 48)
• Calculate the fraction of LRUs hit at k≥12 subcarrier pairs (25%)
Reuse 1, NPRU = 48
LDRU = 6
LDRU = 12
LDRU = 24
LDRU = 48
Proposals
Intel
0.1650
0.0815
0.0398
0.0190
Samsung
0.3026
0.1576
0.0838
0.0484
LGE
0.4753
0.0841
0.0385
0.0138
Conclusions of Part III
• For small number of DRU Intel’s subcarrier
permutation has the best performance.
• Intel and LGE’s subcarrier permutations
have the best performance for large
number of DRUs.
Recommendation
• Adopt the formulas proposed by Intel in
IEEE C802.16m-08/1508r1