Starvation mitigation and priority aware of CSMA/CA in WSN with implementing Markov chain model

Abstract

In the domain of wireless networking, the Carrier Sense Multiple Access with Collision Avoidance (CSMA/CA) has a wide range of applications. Based on CSMA/CA, the IEEE 802.15.6 MAC protocol is used for the body network area. For reduction of the collision’s probability, the accessing the medium, the CSMA/CA uses a distributed coordination function. For a node to transmit the data in CSMA/CA, the channel is sensed and then the transmission is initiated if it happens to be idle. For a contention period, the node is deferred for transmission, if the channel happens to be busy. When the multiple nodes attempt to access a channel, their issue of contention will be solved by this mechanism. As a side effect, there is starvation of the decaying nodes, and the data latency is noticeably increased. For a delay-sensitive application, CSMA/CA protocol becomes inapplicable as the higher risk of fatality is associated with the delay critical applications, such as vehicle cohesion avoidance system and the health-care system.In the modeling process, the channel load is not taken into account at first. For high-priority services, channel detection is performed once. For low priority services, channel detection is performed twice. Markov chain model is used to analyze the influence of algorithm parameter setting on network performance, and it shows great efficiency. Accidents may become inevitable with the exponential latency in the receiving of an emergency notification. Hence, for backing up the loopholes of IEEE 802.15.6 MAC protocol and uprising the response rate of the emergency and warning cases, the author substitutes back off the procedure by proposing a solution that considers the level of emergency or the priority of the message with the help of associated contention window and level of emergency. This reduces the chances of starvation as well. The techniques for CSMA/CA that are priority aware, surely outperform the state of the art back off and this is proved by simulation that the model has enhanced the efficiency of the channel in regards to throughput, delays in transmission, utilization of the bandwidth.