An Efficient MAC Protocol with Correlated Connection Arrival and Variable Slot Assignment in Wireless Sensor Networks

摘要

We present a novel analytical framework to investigate the performance of activity based TDMA slot scheduling under varying traffic conditions in cluster-based wireless sensor networks. Dynamic TDMA Slot Scheduling (DTSS) scheme is presented to analyze the channel utilization and packet dropping probability in bursty network environment under low and high traffic conditions with a large number of sensor nodes in a single cluster. DTSS scheme will allow the network to adapt to the changing traffic load. Based on the network activity, the connection is established between the cluster-head node and those sensors nodes which have data to send, and requested number of TDMA slots is assigned to them dynamically. We have modeled the data traffic by a correlated stochastic process (i.e. Markov chain) for a network where the second and subsequent connection arrival rate is dependent on the first arrival rate. We numerically compared DTSS with traditionally proposed static TDMA model and proved substantial improvement in the bandwidth efficiency and decrease in connection blocking probability in low activity networks. The drawback of the proposed system is also investigated. Results show that the efficiency of the MAC protocol can be increased significantly using the proposed model with reasonably low packet dropping probability.