Proactive Traffic-Adaptive Tuning of Contention Window for Wireless Sensor Network Medium-Access Control Protocol
Pirborj, Nesae Mouzehkesh (2009) Proactive Traffic-Adaptive Tuning of Contention Window for Wireless Sensor Network Medium-Access Control Protocol. Masters thesis, Universiti Putra Malaysia.
The ongoing advances in wireless networks have further expanded the boundaries to the new and challenging area of Wireless Sensor Networks (WSN). Unique properties of sensor nodes such as limited energy storage, constrained processing capabilities and the especially different environments they are usually deployed in have prompted the need of novel protocols in all the layers of the communication stack. A Medium Access Control (MAC) protocol is responsible to sufficiently provide access to a shared medium. Therefore effective techniques in order to reduce the probability of collisions while contending for the medium can be established in a MAC protocol for it organizes the specific time slot a node can have access to the channel. The need for further improving the current applied MAC protocols for WSN in order to reduce the probability of collisions while being energy aware has motivated this research. Sensor MAC as the very first MAC protocol for WSN has been designed on top of the IEEE 802.11 MAC protocol along with some added features to meet the special requirements of a WSN. However the Back-Off scheme of Sensor MAC (S-MAC) is based on a fixed Contention Window (CW) size. This is known as a significant trouble spot in S-MAC in the sense that the delay produced during collisions and idle listening can be so critical to the limited battery lifetime of a sensor node. IEEE 802.11 MAC protocol follows a static approach for obtaining the back-off time and resets the CW to its default minimum upon just one successful transmission and doubles it each time it faces a collision. While the back-off algorithm of IEEE 802.11 suffers from unfairness for its faulty behaviour in both high and low traffic loads the back-off mechanism in S-MAC suffers from a fixed CW size. Reducing the undesired idle listening time caused by unnecessary long back-off times when traffic is low and also decreasing the probability of collisions in situations with high traffic load due to the fixed CW size in S-MAC have motivated our research. We have tried to come up with a dynamic back-off algorithm for SMAC that can extract the current traffic information of the network and engage them in estimating the contention window from which the back-off time is chosen. Our approach is a proactive algorithm to get the CW of the neighbouring nodes ready before contending for the medium. The performance of our algorithm has been measured in terms of average delay, average throughput, delivery ratio, and average energy efficiency. It is shown that our back-off scheme has reduced the delay by 47% and has decreased the energy consumption up to above 15% over the current SMAC implementation. The delivery ratio and throughput have been improved up to 44% and 28% respectively.
Repository Staff Only: Edit item detail