View Full Document

12 views

Unformatted text preview:

Robust Rate Adaptation for 802 11 Wireless Networks Starsky H Y Wong1 Hao Yang2 Songwu Lu1 and Vaduvur Bharghavan3 Dept of Computer Science UCLA 4732 Boelter Hall Los Angeles CA 90025 1 IBM T J Watson Research 19 Skyline Drive Hawthorne NY 10532 2 Meru Networks 1309 South Mary Avenue Sunnyvale CA 94087 3 hywong1 slu cs ucla edu1 haoyang us ibm com2 bharghav merunetworks com3 ABSTRACT Rate adaptation is a mechanism unspecified by the 802 11 standards yet critical to the system performance by exploiting the multi rate capability at the physical layer In this paper we conduct a systematic and experimental study on rate adaptation over 802 11 wireless networks Our main contributions are two fold First we critique five design guidelines adopted by most existing algorithms Our study reveals that these seemingly correct guidelines can be misleading in practice thus incur significant performance penalty in certain scenarios The fundamental challenge is that rate adaptation must accurately estimate the channel condition despite the presence of various dynamics caused by fading mobility and hidden terminals Second we design and implement a new Robust Rate Adaptation Algorithm RRAA that addresses the above challenge RRAA uses short term loss ratio to opportunistically guide its rate change decisions and an adaptive RTS filter to prevent collision losses from triggering rate decrease Our extensive experiments have shown that RRAA outperforms three well known rate adaptation solutions ARF AARF and SampleRate in all tested scenarios with throughput improvement up to 143 Categories and Subject Descriptors C 2 1 ComputerCommunication Networks Network Architecture and Design Wireless communication General Terms Design Experimentation Performance Keywords Rate Adaptation 802 11 1 INTRODUCTION Rate adaptation is a link layer mechanism critical to the system performance in IEEE 802 11 based wireless networks yet left unspecified by the 802 11 standards The current 802 11 specifications mandate multiple transmission rates at the physical layer PHY that use different modulation and coding schemes For example the 802 11b PHY supports four transmission rates 1 11 Mbps the 802 11a PHY offers eight rates 6 54Mbps and the 802 11g PHY sup Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page To copy otherwise to republish to post on servers or to redistribute to lists requires prior specific permission and or a fee MobiCom 06 September 23 26 2006 Los Angeles California USA Copyright 2006 ACM 1 59593 286 0 06 0009 5 00 ports twelve rates 1 54Mbps To exploit such multi rate capability a sender must select the best transmission rate and dynamically adapt its decision to the time varying and location dependent channel quality without explicit information feedback from the receiver Such an operation is known as rate adaptation Given the large numerical span among the available rate options rate adaptation plays a critical role on the overall system performance in 802 11based wireless networks such as the widely deployed WLANs and the emerging mesh networks In recent years a number of algorithms for rate adaptation 1 2 12 8 3 4 10 5 6 7 9 have been proposed in the literature and some 1 12 8 have been used in real products Their basic idea is to estimate the channel quality and adjust the transmission rate accordingly This is typically achieved by using a few metrics collected at the sender and the associated design rules The widely used metrics include probe packets 1 2 8 consecutive successes losses 1 2 6 PHY metrics such as SNR 4 3 6 and long term statistics 12 Examples of the commonly used rules include increasing rate upon consecutive successes using probe packets to assess new rates etc While all such metrics and rules seem intuitively correct and each design has its own merits little is known about how effectively they perform in a practical setting The fundamental problem is that real world wireless networks exhibit rich channel dynamics including random channel errors mobility induced channel variation and contention from hidden stations Each of the above metrics and associated design rules has limited applicable scenarios Consequently each design has its own Achille s heel In this paper we conduct a systematic and experimental study to expose the challenges for rate adaptation and explore new design space To this end we first use experiments and simple analysis to critically examine five design guidelines followed by most existing algorithms These guidelines include 1 decrease transmission rate upon severe packet loss 2 use probe packets to assess the new rate 3 use consecutive transmission successes losses to decide rate increase decrease 4 use PHY metrics to infer new transmission rate and 5 long term smoothened operation produces best average performance For experimental comparison we implement three popular algorithms ARF 1 AARF 2 SampleRate 8 on a programmable AP platform together with the ONOE algorithm 12 available in MADWiFi 17 We not only identify the issues with these algorithms using experiments but also take a microscopic view of their runtime behavior and gain insights on the root causes of the issues Our experiments surprisingly show that the above AP P2 H P4 P3 P1 R P5 Figure 1 Experimental floor plan five seemingly valid guidelines can be quite misleading in practice and may incur significant performance penalty of up to 70 throughput drop In fact we even discovered that with mild link layer contention these rate adaptation designs not only fail to facilitate throughput improvement but also reduce the throughput and aggravate channel contention because rate decrease is falsely triggered To address these challenges we design and implement a Robust Rate Adaptation Algorithm RRAA based on two novel ideas First we use short term loss ratio in a window of tens of frames to opportunistically guide the rate selection Such a loss ratio provides not only fresh but also dependable information to estimate the channel quality Second we leverage the per frame RTS option in the 802 11 standards and use an adaptive RTS filter to suppress collision losses with minimal overhead We implement RRAA on a programmable AP platform and evaluate its performance using thorough experiments


Access the best Study Guides, Lecture Notes and Practice Exams

Loading Unlocking...
Login

Join to view Wong.Mobicom06 and access 3M+ class-specific study document.

or
We will never post anything without your permission.
Don't have an account?
Sign Up

Join to view Wong.Mobicom06 and access 3M+ class-specific study document.

or

By creating an account you agree to our Privacy Policy and Terms Of Use

Already a member?