Effects of Contention Window and Packet Size on the Energy Efficiency of

Wireless Local Area Network

Xiaodong Wang, Jun Yin andDharma P. Agrawal

University of Cincinnati, Cincinnati, Ohio.

IEEE WCNC 2005

Outlines

• Introduction

• Analysis methods– Packet transmission probability– Energy efficiency analysis– Packet error probability

• Simulations

• Conclusions

Introduction

• Technology in battery capacity hasn’t a big advance in recent years

• Energy efficiency is one of the most important challenging problems in wireless communications

Introduction (cont.)

• 802.11 standard specifies a power saving mode (PSM)– Synchronized

• This paper concentrate on 802.11 DCF protocol

Introduction (cont.)

• In 802.11 DCF, energy could be consumed by– Idle listening– Transmission / Receiving

• Packet size

– Collision• Exponential backoff procedure (CW)

Goal

• Optimize energy efficiency by the analytical model– Optimal contention window– Optimal packet size

Packet transmission probability

• The probability to send a packet successfully after i times of unsuccessful transmission is Pi

where pu indicates the unsuccessful probability

• With maximum backoff stage m and retry count m ’

when m’ ≦ m

when m’ ＞ m

Packet transmission probability

• τis the average probability of each node to send a packet if the medium is idle

• Collision probability p

• Transmission error probability

Packet transmission probability

• Probabilities of – transmission in one of the other (n-1) node– contends successfully of any other (n-1) node– transmits successfully of any other (n-1) node

Packet transmission probability

• Average number of unsuccessful transmission

Energy efficiency analysis

Collision

Error

• Time of success / collision

• Total energy consumption of a successful transmission

Energy efficiency analysis

Backoff Freeze Collision Error Success

• Time of backoff period

Energy efficiency analysis

• Average number of transmission overhead can be expressed by

we could get E_FR

Energy efficiency analysis

• E_SU can simply expressed by

• Finally, we get η

Energy efficiency analysis

Simulations

• System parameters

Energy efficiency of DCF under ideal and non-ideal environment

n=20CWmin=32PKT = 1000B

Effects of CWmin on the energy efficiency of DCF

n=20PKT = 1000B

Effects of packet size on the energy efficiency of DCF with RTS/CTS

n=20

Comparison of optimal packet size and optimal CWmin

Conclusions

• This paper present the analysis of the energy efficiency in 802.11 DCF– Compare the impact of CW and Packet

size

• Packet size can effect the energy efficiency under error-prone channel