A channel access cycle based model for IEEE 802.11e EDCA in unsaturated traffic conditions

Abstract

802.11e enhanced distributed channel access (EDCA) introduces different parameters to provide differentiated QoS support for various traffic classes. This brings additional difficulties and complexities to the traditional per-slot based Markov chain modeling techniques, where the transmission probability relies on generic time slots. Channel access cycle based modeling, which does not depend on per-slot states, breaks new ground in 802.11e EDCA study. However, the current cycle based model only considers saturated traffic load conditions, while typical network conditions are unsaturated with significant idle periods. We define a new unsaturated channel access cycle that includes a potential idle period and post-backoff procedure, in addition to the backoff procedure and a transmission attempt of the saturated cycle. A new EDCA model for unsaturated traffic conditions is proposed based on this channel access cycle. Our model achieves high accuracy in describing channel access operations over a wide range of traffic conditions. The model solutions, validated by ns-2 simulations, accurately characterize the unique features of 802.11e EDCA in unsaturated traffic conditions.