Autonomous Clustering and Message Passing Protocol for Energy Efficiency in Wireless Sensor Networks

Wireless sensor networks consist of a large number of nodes with limited battery power and sensing components, which can be used for sensing specified events and gather wanted or interesting information via wireless links. It will enable the reliable monitoring of a variety of environments for both civil and military applications. There is a need of energy-efficient message collection and power management methods to prolong the lifetime of the sensor network. Many methods, such as clustering algorithm, are investigated for power saving reason, however, they only consider reducing the amount of message deliveries by clustering but not the load balance of the clusters to extend the maximum lifetime of the network. Therefore, in this paper, we propose a fully distributed, randomized, and adaptable clustering mechanism named autonomous clustering and message passing (ACMP) protocol for improving energy efficiency in wireless sensor networks. Sensor nodes, according to ACMP, can cluster themselves autonomously by their remaining energy and dynamically choose a corresponding cluster head (CH) to transfer the collected information. Sensor nodes adjust an appropriate power level to form clusters and use minimum energy to exchange messages. The network topology is changed dynamically depending on the CH's energy. Moreover, by maintaining the remaining energy of each node, the traffic load is distributed to all nodes and thus prolong the network lifetime efficiently. Simulation results show that ACMP can achieve a highly energy saving effect as well as prolong the network lifetime.