Abstract

Without a doubt, Wireless Local Area Networks (WLANs) technology has been encountering an explosive growth lately. IEEE 802.11 is the standard associated with this promising technology, which enures shared access to the wireless medium through the distributed coordination function (DCF). Recently, the IEEE 802.11e task group has made extensions to WLANs medium access control (MAC) in order to support quality of service (QoS) traffic. An inherited problem for WLANs, is the volatility of the propagation medium, which is a challenging issue that affects the system performance significantly. Consequently, enhancing the operation of the DCF in noisy environments is of great interest, and has attracted the attention of many researchers. Our first major contribution in the presented thesis, is an analytical and simulation analysis for the binary exponential backoff (BEB) scheme of the DCF, in the presence of channel noise. We show that following the BEB procedure when a host encounters erroneous transmission is needed only if the channel was highly loaded. However, incrementing the contention window (CW) upon each packet failure, whether caused by instantaneous transmission (i.e. collision) or channel noise, will result in the waste of air time if the channel was lightly loaded. Accordingly, we present a hybrid access method that adapts the CW according to the channel load along with the frame error rate (FER). Other means to overcome the channel noise is the adjustment of the transmission rate. Many rate adaptation (RA) algorithms were introduced in the past few years, including the Automatic Rate Fallback (ARF) which is currently implemented in the wireless cards. Yet, many drawbacks are associated with these RA algorithms; specifically, in regard to the techniques and events that should trigger the rate change. Moreover, the IEEE 802.11e QoS flows requirements were not considered with the latter schemes. Accordingly, our next major contribution in this work is the presentation of a novel rate adaptation scheme. The simplicity of the introduced rate adaptation scheme is that it relies on the MAC layer parameters rather than those of the PHY layer when adjusting the rate. Furthermore, our algorithm supports the IEEE 802.11e MAC extensions where QoS traffic requirements were integrated in the procedure of adjusting the bit rate. Hence, strict real-time flow parameters such as delay and maximum drop rate are respected. Finally, we enhance the dynamic assignment of transmission opportunities (TXOPs) in order to offer fair air-time for nodes facing high packet loss rate