D2D: Delay-Aware Distributed Dynamic Adaptation of Contention Window in Wireless Networks

Abstract

The IEEE 802.11e enhanced distributed channel access (EDCA) protocol follows class-based service differentiation for providing differentiated quality-of-service (QoS). However, its collision avoidance mechanism using backoff algorithm can be inefficient for providing improved performance with respect to throughput and channel access delay, especially in a high network configuration (i.e. number of stations) with imperfect wireless channel. The existing and emerging works have devoted considerable attention on tuning the backoff parameters for achieving optimal throughput only. The prior works do not consider the channel access delay and throughput metrics altogether for performance improvement. Additionally, in most of the cases, the optimal configuration of backoff parameters are performed by a centralized controller. In this paper, we propose a delay-aware distributed dynamic adaptation of contention window scheme, namely D2D , for the cumulative improvement of both the throughput and the channel access delay at runtime. The D2D scheme requires two ad-hoc, distributed, and easy-to-obtain estimates— delay deviation ratio and channel busyness ratio —of the present delay level and channel congestion status of the network, respectively. A key advantage of the D2D scheme is that it is compliant with the IEEE 802.11 standard, and, thus, can be seamlessly integrable with the existing wireless card. We show the integrated model of the medium access control protocol, namely D2D Channel Access ( D2DCA ), for the IEEE 802.11e networks. We further propose a two-dimensional Markov chain model of the D2DCA protocol for analyzing its theoretical performance in saturated networks with imperfect wireless channel. Theoretical comparison with the benchmark protocols establishes the effectiveness of the D2DCA protocol.