Improved Moth Flame Optimization Based Cluster Head Selection and Data Aggregation Using Machine Learning Approach

Wireless Sensor Network is comprised with group of sensor nodes which are utilized in various field of applications. These sensor nodes are comprised with minimal processing ability due to diminished battery power. Since WSNs are vulnerable to failure because of power issues, the data aggregation plays a major role in WSN. The majority of the power is wasted due to redundant data from sensor nodes to base station. So, this research introduced an effective data aggregation scheme using Support Vector Machine (SVM) and the selection of CH takes place using the proposed Improved Moth Flame Optimization Algorithm (IMFO). The experimental results show that energy consumption of proposed IMFO for 200 nodes is 89.56 J whereas the existing Cluster based Reliable Data aggregation CRDA consumed 95.45 J. Similarly, the PDR of the proposed approach for 20 nodes is 0.83% whereas the existing Sail Fish Optimization with Support Vector Machine (SFO-SVM) achieved throughput of 0.71%.