Minimizing Energy Consumption Based on Neural Network in Clustered Wireless Sensor Networks
Wireless sensor networks were organized with the collections of sensor nodes for the purpose of monitoring physical phenomenon such as temperature, humidity and seismic events, etc., in the real world environments where the manual human access is not possible. The major tasks of this type of networks are to route the information to sink systems in the sensor network from sensor nodes. Sensors are deployed in a large geographical area where human cannot enter such as volcanic eruption or under the deep sea. Hence sensors are not rechargeable and limited with battery backup; it is very complicated to provide the continuous service of sending information to sink systems from sensor nodes. To overcome the drawback of limited battery power, this paper proposes the concept of minimizing energy consumption with the help of neural networks. The modified form of HRP protocol called energy efficient HRP protocol has been implemented in this paper. Based on this concept, the workload of cluster head is shared by the cluster isolation node in order to increase the lifetime of the cluster head node. Also cluster monitoring node is introduced to reduce the re-clustering process. The implementation procedure, algorithm, results and conclusions were proved that the proposed concept is better than the existing protocols.