Cluster Based Optimization of Routing in Distributed Sensor Networks Using Bayesian Networks with Tabu Search

Distributed Sensor Networks (DSNs) have attracted significant attention over the past few years. A growing list of many applications can employ DSNs for increased effectiveness especially in hostile and remote areas. In all applications a large number of sensors are expected and requiring careful architecture and management of the network. Grouping nodes into clusters has been the most popular approach for support scalability in DSN. This paper proposes a cluster based optimization of routing in DSN by employing a Bayesian network (BN) with Tabu search (TS) approach. BN based approach is used to select efficient cluster heads and construction of BN for the proposed scheme. This approach incorporates energy level of each node, bandwidth and link efficiency. The optimization of routing is considered as a design issue in DSN due to lack of energy consumption, delay and maximum time required for data transmission between source nodes (cluster heads) to sink node. In this work optimization of routing takes place through cluster head nodes by using TS. Simulations have been conducted to compare the performance of the proposed approach with LEACH protocol. The objective of the proposed work is to improve the performance of network in terms of energy consumption, throughput, packet delivery ratio, and time efficiency of optimization of routing. The result shows that the proposed approach perform better than LEACH protocol that utilizes minimum energy, latency for cluster formation and reduce overhead of the protocol.