Quality of Service Provisioning in Wireless LANs

Featuring Inanc Inan, Ph.D.
Research Scientist
Wionics Technologies
Realtek Semiconductor Corp.

Location:  Engineering Gateway 3161
Free and open to the public

Abstract: 

Recent years have witnessed a remarkable growth of Wireless Local Area Networks (WLANs). The IEEE 802.11 standard specifies the Physical (PHY) layer and the Medium Access Control (MAC) layer algorithms for  WLANs. Due to the specific design of MAC layer algorithms, IEEE 802.11 WLANs can only provide best effort services and do not have Quality-of-Service (QoS) support for multimedia applications. Therefore, recently, a new QoS-enhanced standard, IEEE 802.11e, has been developed. The IEEE 802.11e standard introduces the Hybrid Coordination Function (HCF) containing two medium access mechanisms: a contention-based channel access, namely Enhanced Distributed Channel Access (EDCA), and a polling-based channel access, namely HCF Controlled Channel Access (HCCA). 

In this talk, we propose various analytical models for accurate performance analysis of the EDCA function. As we discuss, the proposed  analytical frameworks can also address the multimedia capacity estimation problem.  Based on this observation, we design simple admission control algorithms that enable comprehensive QoS support in the next-generation WLAN. 

We first present two analytical models for capacity estimation of EDCA in saturation (at arbitrarily high load). The first model we propose  is a Discrete-Time Markov Chain (DTMC) which incorporates an accurate Contention Window (CW) and Arbitration Interframe Space (AIFS) differentiation for any number of active Access Categories (ACs). The second model we propose uses the fact that random access schemes present cyclic behavior. The analysis shows that a station- and AC-specific cycle time exists for the EDCA function. Validating the theoretical results via simulations, we show that both models accurately capture EDCA saturation performance, and are more accurate than the models that miss accurate handling of the protocol details.

Next, we propose an analytical model for capacity estimation of EDCA function which incorporates an accurate CW, AIFS, and Transmit Opportunity (TXOP) differentiation at any traffic load. The proposed model is shown to capture the effect of MAC layer buffer size on the performance. Analytical and simulation results are compared to demonstrate the accuracy of the proposed approach for varying traffic loads, EDCA parameters, and MAC layer buffer space. The model can directly be employed in an EDCA admission control algorithm.

For multimedia capacity analysis of the EDCA function, we also propose another analytical framework. This framework utilizes the results of saturation analysis in the capacity estimation. A simple admission control method is proposed. We show that the proposed admission control algorithm maintains satisfactory user-perceived quality for coexisting voice and video connections in an infrastructure Basic Service Set (BSS) and does not present over- or under-admissionproblems of previously proposed models in the literature.

About the Speaker:

Inanc Inan received the B.S. degree from the Middle East Technical University, Ankara, Turkey, in June 2001, an M.S. degree from Bilkent University, Ankara, Turkey, in September 2003, and a Ph.D. degree from University of California, Irvine, in September 2007, all in electrical engineering.  He was a CPCC Fellow during 2004-2007. He was with Wireless Networking division of Conexant Systems, Inc., during Summer 2006.

Since July 2007,  he has been working as a research scientist at Wionics Research - Realtek Group, concentrating mainly on MAC, link, and transport layer protocol esearch and development for UWB wireless networks. His current research interests include analytical network modeling and simulation, QoS provisioning, fair and efficient resource allocation, protocol and algorithm design for wireless networks.