Performance analysis of AMCLM protocol (adaptive multi-services cross-layer MAC protocol)

Abstract

Listen

Translate article

The weakness of radio transmission is due to the variation of link quality which is disturbed by noise interference, multipath free space loss and multipath. In Wifi MAC protocols, devices don't take into account the radio link transmission quality when they decide to transmit. In a previous study, we have pro- posed a new MAC protocol which is named AMCLM (Adaptive Multi-services Cross-Layer MAC protocol). It CONSIDERS PHY layer's constraints. Thus, if the node transmission channel SNR (Signal to Noise Ratio) is under specified threshold, the protocol decides to disassociate the node in order to improve quality of service of other connected nodes in the BSS (Basic Service Set). In this study we make performance analysis of this protocol using analytical method based on discrete time markov chains. Keywords: 802.11, QoS, Cross-layer, PHY, MAC, SNR (Signal to Noise Ratio), Interlacing, Makov chain, Performance evaluation. I. INTRODUCTION The CSMA/CA MAC protocol is developed in order to share bandwidth in efficient manner between all users con- nected to an access point (BSS Basic Service Set) in WIFI net- works. This protocol is an adaptation of CSMA/CD protocol in wireless networks which avoid a collision. However, it does not include the radio link vulnerability attributed to the effects such as noise, interference, free-space loss, shadowing and multipath fading. In (7) they show that the maximal throughput always falls to the minimal one of all nodes and that the effective bandwidth is the available one divided on the mobile nodes number. Based on this last point we have developed in (2) a new technique which increases the quality of service of all nodes in the BSS by disassociating a node if the SNR is very low. The originality of our approach is that it can be easily designed in existing systems only by adding software in devices. In this study we make performance evaluation of our technique using analytical method based on discrete Markov Chain resolution. We will discuss in next section related works to our field of study. Then after a brief description of our new protocol, we will give an overview of our mathematical model, finally we will summarize the main contribution of this work.