Optimized Query Response Ordering Data Aggregation Models in Wireless Sensor Network

Abstract

Wireless Sensor Networks are witnessed in both academia and industry as an emerging and exciting area. The self-management sensor network has certain technological problems, such as quick depletion of energy, limited network bandwidth. Application- specific installation makes it more challenging for effective utilization. Our contribution focuses on an enhanced network performance by implementing various query response order-based data aggregation models with the inculcation of required optimization techniques and neural network models. The research contribution focuses on three significant contributions; First, we create an aggregation model to order the responses of sensor nodes to the base station. The efficient ordering leads to optimal performance and we thus consider the data aggregation model of the encoded query response order. The proposed model is then designed to rank the latency and throughput based on Query Response Order (QRO) using the superior Group Search Optimiza- tion (GSO) optimization algorithm. Therefore, The efficiency of data aggregation is enhanced against the conventional approaches using the predicted GSO-based Query Response Order by reaching peak throughput and minimal latency. Secondly, we integrate the neural network with the GSO algorithm to make the query response-based data aggregation model self-regulating and knowledgeable. In the proposed dual-stage optimization process, we produce a library of optimized query response ordering using a Neural network for preparation, while the second stage is used as an input to the GSO algorithm to obtain refined Query response order for data aggregation effectiveness. Finally, the work shelters an essential change in the Lion algorithm in the mating phase and called it as Fitness-Mated Lion Algorithm (FM-LA). The multi-objective function covers essential parameters such as data freshness, latency, and network throughput. Initially, query response order is generated using FM-LA given as an input to Neural Network (NN). Thus, the optimal qu