An Improved Analytical Model of 802.11e EDCA with Variable Packet Length
G Wang,SL Mei,R Yu,ZY Zhang
DOI: https://doi.org/10.1109/vetecf.2005.1558101
2005-01-01
Abstract:An improved two-dimensional analytical model of 802.11e enhanced distributed channel access (EDCA) is pre- sented. The model considers the node transmission probability to investigate the impact of internal collision resolution (ICR) mechanism besides parameters differentiation, which is simpler than the three-dimensional Markov chain analysis. An additional feature is that it introduces a novel approach to find the normalized throughput and packet delay of traffics when they have different packet length. The simulation results validate the analysis and demonstrate the impact of ICR and traffic packet length. I. INTRODUCTION The IEEE 802.11 has gained great popularity in the de- ployment of wireless local area network (WLAN). In the MAC layer, the standard defines two forms of medium access, known as distributed coordination function (DCF) and point coordination function (PCF). However, These mechanisms only provide limited Quality of Service (QoS) support for various kinds of traffic, especially those with stringent real- time requirements. In order to provide a better QoS support for different traffic requirements, IEEE 802.11e (1) has presented a new channel access function, called hybrid coordination function (HCF). The HCF includes a contention-based channel access referred to as the enhanced distributed channel access (EDCA) and a HCF controlled channel access called HCCA. Due to the easy deployment nature of EDCA, in this paper, we confine our analysis to the EDCA scheme. In this media access mechanism, traffics with various priorities are mapped into different access categories (ACs) to contend the channel. On the condition that more than one AC inside a node finishes backoff at the same time, the internal collision resolution (ICR) mechanism will pick up the highest priority traffic to transmit and backoff the others. Intuitively, the ICR provides more priority than the defined channel contention parameters. The model analysis of EDCA has been a research focus since the IEEE 802.11e draft was proposed. Most of the work can be regarded as an extension to an IEEE 802.11 DCF analytical model provided by (2). It used a two-dimensional Markov chain to analyze the saturation throughput of IEEE 802.11 DCF and got a perfect result. Following (2), (3) pro- posed an improved DCF model which considered the backoff suspension during busy time slots. The model in (4) analyzed the parameters differentiation provided in EDCA while ig- nored the impact of ICR. (5) and (6) proposed two three- dimensional Markov models to analyze the service differen- tiation in EDCA, which increased the analysis complexity. Moreover, the EDCA models mentioned above are analyzed on the assumption that all kinds of traffic have the same packet length. However, This is likely to expose deficiencies in the model analysis because according to different traffic characteristics, there must be mixed packets with various length on the channel. In this paper, we propose an improved two-dimensional analytical model of IEEE 802.11e EDCA. It introduces node transmission probability to investigate the impact of ICR. Meanwhile, the model under various traffic conditions is analyzed. And the simulation results validate our analysis. The paper is outlined as follows. In section II we briefly review the media access mechanism of IEEE 802.11e EDCA. In section III we propose a simple two-dimensional Markov analytical model and consider the impact of different packet length at the same time. The validation of the proposed model is provided in Section IV. Section V presents the concluding remarks.