Scale WLANs Utilizing a Power

blessinghomoeopathAI and Robotics

Nov 30, 2013 (3 years and 6 months ago)

94 views

5/2/2007

IEEE Sarnoff Symposium 2007

1

Optimal Load Distribution in Large
Scale WLANs Utilizing a Power
Management Algorithm

Presented by Mohamad Haidar, Ph.D.
Candidate

University of Arkansas at Little
Rock


5/2/2007

IEEE Sarnoff Symposium 2007

2

Presentation Outline


Introduction


Wireless Local Area Networks (WLANs)


Access Points (APs) congestion


Problem Statement


Related Work


Proposed Solution


Minimizing the load at the Most Congested AP (MCAP)


Power Management Algorithm


Problem Formulation


Algorithm


Numerical Analysis


Results


Conclusion


Future work

5/2/2007

IEEE Sarnoff Symposium 2007

3

Introduction


Wireless Local Area
Networks.


Airports


Hotels


Colleges



What is AP congestion?

C
AP
= (R
1
+ R
2+..+
R
N
)/BW


C
AP
: Congestion at AP

R : Data rate of a user connected to the
AP

BW: Bandwidth (i.e. BW=11 Mbps for
IEEE 802.11b)

1 2
...
j
N
j
U U U
BW
  
5/2/2007

IEEE Sarnoff Symposium 2007

4

Problem Statement


Congestion at hot spots



Degrades network throughput


Slowest station will make other stations wait
longer


Unfair load distribution over the network
causes bottlenecks at hot spots.


Inefficient bandwidth utilization of the
network



5/2/2007

IEEE Sarnoff Symposium 2007

5

Related Work

1.
Y. Lee, K. Kim, and Y. Choi. “Optimization of AP placement and Channel Assignment in
Wireless LANs.”

LCN 2002. 27th Annual IEEE Conference on Local Computer Networks,

pages 831
-
836, November 2002.


ILP


2.
R. Akl and S. Park. “Optimal Access Point selection and Traffic Allocation in IEEE
802.11 Networks,”

Proceedings of 9th World Multiconference on Systemics, Cybernetics
and Informatics (WMSCI 2005): Communication and Network Systems, Technologies and
Applications
, paper no. S464ID, July 2005.


ILP


3.
I. Papanikos, M. Logothetis, "A Study on Dynamic Load Balance for IEEE 802.11b
Wireless LAN,"
Proc. 8th

International Conference on Advances in Communication &
Control,

COMCON 8
, Rethymna, Crete/Greece, June 2001.



Number of users associated rather than traffic load.


4.
H. Velayos, V. Aleo, and Karlsson, “Load Balancing in Overlapping Wireless LAN Cells”,
Proceedings of IEEE ICC 2004
, Paris, France, June 2004.


Balance index.







5/2/2007

IEEE Sarnoff Symposium 2007

6

Proposed Solution


We propose solving the congestion at the hot spots
by decrementing the power transmitted by the
MCAP in discrete steps until one or more users can
no longer associate with any AP or their data rate
can no longer be accommodated.


Advantages:


Load is fairly distributed


Increase in data rate throughput per user


Less adjacent and co
-
channel interference.



5/2/2007

IEEE Sarnoff Symposium 2007

7

Problem Formulation


MCAP ILP formulation:




Subject to


min
ij
x
1 2
max{,,...,}
M
C C C
1
i M
 
1
j N
 
1
1
N
ij
i
x



1
M
i ij
j
j
U x
Cj
BW




for
j= 1,…, M

for
i= 1,…,N

5/2/2007

IEEE Sarnoff Symposium 2007

8

Algorithm


Compute Received Signal Strength Indicator (RSSI) at
each user


Generate a binary matrix that assigns “1” if a user’s
RSSI exceeds the threshold value or “0” otherwise.


Invoke LINGO to solve the ILP


Identify the MCAP


Decrement its transmitted power by 1 dbm


Repeat previous steps until one or more user can no
longer associate with an AP.


Observe the power levels at each AP and the best
user’s association and best loads at APs.


5/2/2007

IEEE Sarnoff Symposium 2007

9

Numerical Analysis


Receiver Sensitivity at the
user is
-
90 dBm


Transmitted Power at each
AP is 20 dbm

User Number

AP1

AP2

AP3

AP4

1

1

1

1

0

2

0

0

1

0

3

0

0

0

1

4

0

0

0

1

5

1

1

1

1

6

1

1

0

0

7

0

1

0

0

8

0

0

1

1

9

0

0

1

0

10

0

1

0

1

4

1

2

5/2/2007

IEEE Sarnoff Symposium 2007

10

Numerical Analysis Continued


Traffic is randomly generated
between 100Kbps and 1Mbps for
each user

User Number

Traffic
(Kbps)

1

741

2

566

3

667

4

467

5

576

6

349

7

738

8

936

9

683

10

805

Service Area Map

5/2/2007

IEEE Sarnoff Symposium 2007

11

Results

User
Number

AP1

AP2

AP3

AP4

1

0

1

0

0

2

0

0

1

0

3

0

0

0

1

4

0

0

0

1

5

0

1

0

0

6

1

0

0

0

7

0

1

0

0

8

0

0

0

1

9

0

0

1

0

10

0

1

0

0


Each user is associated to
one and
ONLY

one AP.

1

1

1

5/2/2007

IEEE Sarnoff Symposium 2007

12

Results Continued


Load is distributed fairly among APs.


Final transmitted power levels at each AP is: 11
dBm, 9 dBm, 4 dBm and 3 dBm, respectively.


Initial Congestion
factor:

(No Power Mgmt)

Congestion factor
solution
according to [2]

Congestion factor
with Power
Mgmt

AP1


0.7323


0.5416


0.4404

AP2


0.4735


0.5378


0.4155

AP3


0.2283


0.3547


0.4559

AP4


0.2393


0.2393


0.3615

Congestion Factor comparison

5/2/2007

IEEE Sarnoff Symposium 2007

13

Results Continued


Different radii sizes
after power
adjustment


Users do NOT
always associate to
the closest AP.

Service area map after Power Mgmt

5/2/2007

IEEE Sarnoff Symposium 2007

14

Results Continued

4 APs

9 APs

16 APs

5/2/2007

IEEE Sarnoff Symposium 2007

15

Conclusion


We proposed an algorithm to reduce
congestion and distribute the load
fairly among APs while adjusting the
transmitted power level at APs.


The model has shown to perform
well for networks of different
topologies.



5/2/2007

IEEE Sarnoff Symposium 2007

16

Future Work


Work is undergoing to


extend the model to include inter
-

and
intra
-
cell interferences and channel
assignments.


Apply the model to a dynamic user
distribution that changes over time.