channel measurements in corridors ftgfor tgac · measurement result: capacity 2n measurement: nlos...
TRANSCRIPT
May 2009 doc.:IEEE 802.11-09/0542r0
Channel Measurements in Corridorsf TGfor TGac
Name Affiliations Address Phone emailName Affiliations Address Phone emailByung-Jae Kwak ETRI 138 Gajeongno, Yuseong-gu,
Daejeon, 305-700, Korea +82-42-860-6618 [email protected]
Jae Joon Park ETRI 138 Gajeongno, Yuseong-gu,
Daejeon, 305-700, Korea +82-42-860-3958 [email protected]
Myung Don Kim ETRI 138 Gajeongno, Yuseong-gu, +82-42-860-6178 [email protected]
Daejeon, 305-700, Korea Minho Cheong ETRI 138 Gajeongno, Yuseong-gu,
Daejeon, 305-700, Korea +82-42-860-5635 [email protected]
Hyun Kyu Chung ETRI 138 Gajeongno, Yuseong-gu,
Daejeon, 305-700, Korea +82-42-860-6140 [email protected]
Sok-Kyu Lee ETRI 138 Gajeongno, Yuseong-gu, +82-42-860-5919 [email protected] 138 Gajeongno, Yuseong gu, Daejeon, 305-700, Korea
82 42 860 5919 @
Submission Byung-Jae Kwak et al., ETRISlide 1
C t tMay 2009 doc.:IEEE 802.11-09/0542r0
ContentsMeasurement campaignsMeasurement results
Delay SpreadCapacityPower delay profile
Conclusion
Submission Byung-Jae Kwak et al., ETRISlide 2
M t C iMay 2009 doc.:IEEE 802.11-09/0542r0
Measurement Campaigns1st Channel measurements
Measurement siteBldg. #11, ETRI: metal wall, closed ended
Antenna configurationAntenna configurationUniform Linear Array Antenna: 8x84 LOS, 8 NLOS
2nd Ch l t2nd Channel measurementsMeasurement site
Bldg. #1, ETRI: concrete walls with metal doors, open endedBldg. #1, ETRI: concrete walls with metal doors, open endedAntenna configurations
Uniform Linear Array Antenna: 8x8, 4x4, 1x16 LOS 6 NLOS6 LOS, 6 NLOS
100 MHz BW at 5.25 GHz
Submission Byung-Jae Kwak et al., ETRISlide 3
May 2009 doc.:IEEE 802.11-09/0542r0
M t Sit 1 (1/2)M t Sit 1 (1/2)Measurement Site 1 (1/2)Measurement Site 1 (1/2)LOS: 6, 7, 8, 9, , ,NLOS: 1, 2, 3, 4, 5, a, b, c
c
desks
6789Tx
10msteel wallconcrete wall
21 3 4 5
a
b
Rx10m
partition
b
Submission Byung-Jae Kwak et al., ETRISlide 4
May 2009 doc.:IEEE 802.11-09/0542r0
M t Sit 1 (2/2)M t Sit 1 (2/2)Measurement Site 1 (2/2)Measurement Site 1 (2/2)Tx antenna height = 2m, Rx antenna height = 1mg , gTwo independent channel measurements at each location (Rx rotated by 90°)
TxTx TxTx
Rx RxRx
Parallel to Tx antenna Perpendicular to Tx antenna
Submission Byung-Jae Kwak et al., ETRISlide 5
May 2009 doc.:IEEE 802.11-09/0542r0
M t Sit 2 (1/2)M t Sit 2 (1/2)Measurement Site 2 (1/2)Measurement Site 2 (1/2)LOS: 1, 2, 3, a, b, c
window
NLOS: 4, 5, 6, 7, 8, 9
6
789
c
10m3.5m
Concretewall
5b
wall
Concrete
4a
Concretewallsteel wall
21 3
Tx
Rx10m
10m
10m
3.5m
Submission Byung-Jae Kwak et al., ETRISlide 6
May 2009 doc.:IEEE 802.11-09/0542r0
M t Sit 2 (2/2)M t Sit 2 (2/2)Measurement Site 2 (2/2)Measurement Site 2 (2/2)Tx antenna height = 2m, Rx antenna height = 1mTwo independent channel measurements at each location (Rx rotated 90°)
Tx
R
Tx
Rx
Rx
Submission Byung-Jae Kwak et al., ETRISlide 7
Parallel to Tx antenna(#1) Perpendicular to Tx antenna(#a)
May 2009 doc.:IEEE 802.11-09/0542r0
M t R lt D l S d1st measurement:
NLOS b t 80 DS LOS 110 DS
Measurement Result: Delay Spread
NLOS: about 80 nsec rms DS; LOS: 110 nsec rms DS
2nd measurement: about 35 nsec rms DS
0 8
0.9
1ETRI Indoor Measurement Results
0.6
0.7
0.8
F
0.3
0.4
0.5
CD
F
0
0.1
0.2 1st Measured-LOS1st Measured-NLOS2nd Measured-LOS2nd Measured-NLOS
Submission Byung-Jae Kwak et al., ETRISlide 8
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
x 10-7
0
Delay Spread [sec]
May 2009 doc.:IEEE 802.11-09/0542r0
M t R lt C it1st measurement: NLOS > LOS
d ( hi l )
Measurement Result: Capacity
2nd measurement: NLOS < LOS ( More on this later)LOS: 1st and 2nd measurements agree with each other
0 8
0.9
1ETRI Indoor Measurement Results
1st Measured-LOS1st Measured-NLOSi.i.d. Channel
0.6
0.7
0.8 2nd Measured-LOS2nd Measured-NLOS
0.3
0.4
0.5
CD
F
0
0.1
0.2
Submission Byung-Jae Kwak et al., ETRISlide 9
15 20 25 30 35 40 45 50 55 600
Capacity [bps/Hz]
May 2009 doc.:IEEE 802.11-09/0542r0
M t R lt M d l GMeasurement Result: Model GModel G (based on Model D)
D l S d M d l D (50 )Delay Spread: same as Model D (50 nsec)Angular Spread: ¼ of Model DAoA: +30o, -30o, -180o, AoD: +30o, -30o, 0o
0.9
1ETRI Indoor Measurement Results
1st Measured-LOS1st Measured-NLOS
AoA: +30 , 30 , 180 , AoD: +30 , 30 , 0
0.6
0.7
0.8i.i.d. Channel2nd Measured-LOS2nd Measured-NLOSModel G LOSModel G NLOS
0 3
0.4
0.5
CD
F
0.1
0.2
0.3
Submission Byung-Jae Kwak et al., ETRISlide 10
15 20 25 30 35 40 45 50 550
Capacity [bps/Hz]
May 2009 doc.:IEEE 802.11-09/0542r0
Measurement Results: Power Delay ProfileMeasurement Results: Power Delay ProfileMeasurement Results: Power Delay ProfileMeasurement Results: Power Delay Profile1st measurement
LOS & NLOS 4 l t2nd measurement
LOS 4 l tLOS & NLOS: 4 clustersNLOS: more rays within clusterNLOS > LOS
LOS: 4 clustersNLOS: 2+ clustersLOS & NLOS: about the same rays within cluster (see 1st cluster)
0Power Delay Profile - 1st Measurement
LOS Position #60
Power Delay Profile - 2nd Measurement
LOS Position #1
within cluster (see 1st cluster)LOS > NLOS
-20
-10NLOS Position #2
-20
-10NLOS Position #6
-40
-30
Pow
er [d
B]
-40
-30
Pow
er [d
B]
-50
-40
-50
-40
Submission Byung-Jae Kwak et al., ETRISlide 11
0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
x 10-6
-60
Delay [sec]0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
x 10-6
-60
Delay [sec]
May 2009 doc.:IEEE 802.11-09/0542r0
Concluding RemarksConcluding RemarksConcluding RemarksConcluding RemarksOur measurement results show that TGn channel models do not model Corridor environment very wellCorridor environment very wellSite dependency of Capacity
LOS capacity is not affected by measurement sitesNLOS capacity varies significantly depending on the measurement
• 1st measurement site: metal walls, closed ended• 2nd measurement site: concrete walls, open ended
Model G: Corridor ModelIn terms of # clusters (3~4), model G is closer to model D than BM d l G h bt i d f d l D b d i th AS b f t 1/4Model G, when obtained from model D by reducing the AS by a factor 1/4, seems to agree with the measurement dataThis is a tentative model, and further work is required
Need more measurement and analysisto verify the above approach, orto develop a new Model G with appropriate parameters (DS AS AoA
Submission Byung-Jae Kwak et al., ETRISlide 12
to develop a new Model G with appropriate parameters (DS, AS, AoA, AoD, etc)