Embed Size (px)
Citation preview
doc.: IEEE 802.15-14-0251-01-0008
Submission
Nah-Oak Song et al.
May 2014
Slide 1
Project: IEEE P802.15 Working Group for Wireless Personal Area Networks (WPANs)
Submission Title: Clustered Random Drop of PDs for Performance Evaluation of PACDate Submitted: May 5, 2014Source: Nah-Oak Song (KAIST), Junhyuk Kim (KAIST), June-Koo Kevin Rhee (KAIST), Byung-Jae Kwak (ETRI), Kapseok Chang (ETRI), Moon-Sik Lee (ETRI)Address: KAIST, Daejeon, Korea; ETRI, Daejeon, KoreaVoice:E-Mail: [email protected], [email protected], [email protected], [email protected], [email protected], [email protected]
Re: TG8 Technical Guidance Document (DCN 15-12-0568-08)
Abstract: This document proposes a new method of randomly distributing PDs for performance evaluation of PAC network. The proposed method produces a more realistic distribution of PDs compared to the conventional uniform random distribution.
Purpose: Discussion.
Notice: This document has been prepared to assist the IEEE P802.15. 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 acknowledges and accepts that this contribution becomes the property of IEEE and may be made publicly available by P802.15.
doc.: IEEE 802.15-14-0251-01-0008
Submission
Nah-Oak Song et al.
Clustered Random Drop of PDs for Per-formance Evaluation of PAC
May 2014
Nah-Oak Song, Junhyuk Kim,
Byung-Jae Kwak, Kapseok Chang
May 2014
Slide 2
doc.: IEEE 802.15-14-0251-01-0008
Submission
Nah-Oak Song et al.
Unique Characteristics of PAC Network
• Infra-less: PDs are not attached to a BS or an AP
• Multi-hop: a single-hop link between PDs is not always guaranteed (hidden terminal prob-lem)
• Mobility: PDs are not expected to stay at the same location
• Clustering: PDs flock around points of attrac-tions rather than points of connections
May 2014
Slide 3
doc.: IEEE 802.15-14-0251-01-0008
Submission
Nah-Oak Song et al.
Uniform Drop of PDs
• Use uniform random distribution to drop (i.e., locate) devices
• Simple but unrealistic– Interference between neighboring devices is underrepre-
sented: not appropriate for D2D, MU-MIMO, etc.– Simulation of (localized) high density of devices is not possi-
ble
• Widely used for SLS of cellular systems– Communication is mostly between BS and UE, and interfer-
ence between neighboring UEs is avoided by scheduling
May 2014
Slide 4
doc.: IEEE 802.15-14-0251-01-0008
Submission
Nah-Oak Song et al.
Miami Map
May 2014
Slide 5
doc.: IEEE 802.15-14-0251-01-0008
Submission
Nah-Oak Song et al.
An Example of Real Device Distribution
May 2014
Slide 6
Android Users in Miami(Source: http://www.businessinsider.com/android-is-for-poor-people-maps-2014-4)
doc.: IEEE 802.15-14-0251-01-0008
Submission
Nah-Oak Song et al.
500m x 500m Area
May 2014
Slide 7
Android Users in Miami(Source: http://www.businessinsider.com/android-is-for-poor-people-maps-2014-4)
doc.: IEEE 802.15-14-0251-01-0008
Submission
Nah-Oak Song et al.
Clustered Random Device Drop• Let be the probability distribution of the devices after
dropping devices, then
where
May 2014
Slide 8
doc.: IEEE 802.15-14-0251-01-0008
Submission
Nah-Oak Song et al.
Evolution of PD Distribution
Slide 9
0 1 2
3 4 5
9 14 19
doc.: IEEE 802.15-14-0251-01-0008
Submission
Nah-Oak Song et al.
Model Parameters
• : The number of PDs
– If , uniform distribution– If , “pull” by already dropped PDs solely de-
termine the location of next drop• : • : bivariate Gaussian with standard deviation
May 2014
Slide 10
doc.: IEEE 802.15-14-0251-01-0008
Submission
Nah-Oak Song et al.
Effect of Parameters
May 2014
Slide 11
𝛽=0.2 𝛽=0.5 𝛽=0.8
𝜎=10
𝜎=20
𝜎=30
0 20 40 60 80 100 120 140 160 180 2000
20
40
60
80
100
120
140
160
180
200
x - position [m]
y -
pos
ition
[m
]User Distribution in the Area of200200[m2] =0.2 and =10[m]
0 20 40 60 80 100 120 140 160 180 2000
20
40
60
80
100
120
140
160
180
200
x - position [m]
y -
pos
ition
[m
]
User Distribution in the Area of200200[m2] =0.5 and =10[m]
0 20 40 60 80 100 120 140 160 180 2000
20
40
60
80
100
120
140
160
180
200
x - position [m]
y -
pos
ition
[m
]
User Distribution in the Area of200200[m2] =0.8 and =10[m]
0 20 40 60 80 100 120 140 160 180 2000
20
40
60
80
100
120
140
160
180
200
x - position [m]
y -
pos
ition
[m
]
User Distribution in the Area of200200[m2] =0.8 and =20[m]
0 20 40 60 80 100 120 140 160 180 2000
20
40
60
80
100
120
140
160
180
200
x - position [m]
y -
pos
ition
[m
]
User Distribution in the Area of200200[m2] =0.8 and =30[m]
0 20 40 60 80 100 120 140 160 180 2000
20
40
60
80
100
120
140
160
180
200
x - position [m]
y -
pos
ition
[m
]
User Distribution in the Area of200200[m2] =0.5 and =30[m]
0 20 40 60 80 100 120 140 160 180 2000
20
40
60
80
100
120
140
160
180
200
x - position [m]
y -
pos
ition
[m
]
User Distribution in the Area of200200[m2] =0.2 and =30[m]
0 20 40 60 80 100 120 140 160 180 2000
20
40
60
80
100
120
140
160
180
200
x - position [m]
y -
pos
ition
[m
]
User Distribution in the Area of200200[m2] =0.2 and =20[m]
0 20 40 60 80 100 120 140 160 180 2000
20
40
60
80
100
120
140
160
180
200
x - position [m]
y -
pos
ition
[m
]
User Distribution in the Area of200200[m2] =0.5 and =20[m]
doc.: IEEE 802.15-14-0251-01-0008
Submission
Nah-Oak Song et al.
Examples of Clustered Random Drop
May 2014
Slide 12
0 50 100 150 200 250 300 350 400 450 5000
50
100
150
200
250
300
350
400
450
500
x - position [m] y
- p
ositi
on [
m]
User Distribution in the Area of500500[m2] =1 and =50[m]
0 50 100 150 200 250 300 350 400 450 5000
50
100
150
200
250
300
350
400
450
500
x - position [m]
y -
pos
ition
[m
]
User Distribution in the Area of500500[m2] =0.6 and =50[m]
0 50 100 150 200 250 300 350 400 450 5000
50
100
150
200
250
300
350
400
450
500
x - position [m]
y -
pos
ition
[m
]
User Distribution in the Area of500500[m2] =0.2 and =50[m]
• 500m x 500m area• 2000 PDs• m
(a) (b) (c)
doc.: IEEE 802.15-14-0251-01-0008
Submission
Nah-Oak Song et al.
Examples of Clustered Random Drop
• Clustered vs. Uniform: # PDs in Range
May 2014
Slide 13
0 100 200 300 400 500 600 700 800 900 1000 11000
0.005
0.01
0.015
0.02
0.025
0.03
0.035
0.04
0.045
Node Degree (Num of PDs within meters).
Sha
re in
the
Are
a
500 500 m2 area, = 50 m, 2,000 PDs
= 0.2 = 0.6 = 1.0(Uniform)
100
101
102
103
0
0.005
0.01
0.015
0.02
0.025
0.03
0.035
0.04
0.045
Node Degree (Num of PDs within meters).
Sha
re in
the
Are
a
500 500 m2 area, = 50 m, 2,000 PDs
= 0.2 = 0.6 = 1.0(Uniform)
doc.: IEEE 802.15-14-0251-01-0008
Submission
Nah-Oak Song et al.
Conclusion
• Unrealistic device drop model can lead to incorrect performance evaluation
• Proposed “Clustered Random Drop” model produces realistic device distribution– Localized high density of clustered PDs is well
represented– Expected to produce more accurate performance
evaluation
May 2014
Slide 14
![IEEE VLCC090917.ppt [互換モード]...... IEEE 802.15- VLC Physical Specification ProtocollayerpositioningProtocol layer positioning & Market Requirements](https://img.pdfslide.tips/doc/110x75/5ad435e07f8b9a571e8beb8b/ieee-ieee-80215-09-0631-00-0007-vlc-physical-specification.jpg)
