Efficient Hybrid Mayfly-Harris Hawks Optimization Support Vector Machine (EMHHO-SVM) Based Data Aggregation and Clustering Technique for Wireless Sensor Networks

Wireless sensor networks encounters a wide variety of applications ranging from disaster management, environment monitoring, smart farming, healthcare, smart cities and mission-critical applications. Wireless sensor node deployed in an area senses the environment and acquires data which gets transferred to the base station (BS) for decision-making and analytics. Since the wireless sensor nodes are resource-constrained in nature, the major portion of the energy is utilized for the transmission of redundant data. This results in the delay of data transmission, high energy consumption and ultimately the decrease in the lifetime of the network. Hence data aggregation and clustering plays a prominent role in alleviating energy consumption in a resource constrained environments. Therefore the problem of energy efficiency can be modeled as an optimization problem. To solve the problem, an energy efficient data aggregation and clustering approach is proposed which combines multi-objective clustering and a nature inspired meta-heuristic optimization routing algorithm. The proposed method involves the use of K-means Clustering for cluster formation. The cluster head election is performed by means of Mayfly algorithm and the data aggregation is performed by considering multiple objectives to reduce energy consumption using SVM. Further, SVM is utilized for Data classification which guarantees the reduction in redundant data transmission. The routing of data from the CHs to the BS is performed using the HHO. The proposed methodology has been implemented for validation using MATLAB R2008a simulation tool. A comparative analysis has been carried out using performance metrics such as delay, packet delivery rate, energy consumption, network lifetime, throughput and packet loss ratio between the existing techniques to that of the proposed model. The methodology has achieved a significant increase in the network lifetime of 32.55%; decrease in energy consumption of 43%; with minimum a low computational overhead of 93%; when compared with its counterparts.