Two-State Hybrid Learning Approaches for Energy Reduction Estimation on Wireless Sensor Networks

Abstract

Wireless Sensor Networks (WSNs) with current technologic advances have been improvised with prime development scenarios that would impact the different areas of research in computer science. Several researchers have been providing real-time solutions with data-mining strategies in wireless environments where these sensor models tend to implement the lifetime of the network and also to reduce energy usage. This paper presents a design model for the WSN network based on optimized link states modeling to identify the specific nodes where the energy minimum has been established. The generated energy estimations where each set of established relay points ensures the energy minimum improvising the parametric criteria on energy will be dealt with the design factors that govern the learning model. The LEACH-A and LEACH-D are implemented with a Fuzzy logic model for the energy reduction criteria. The LSR-MSR protocol where node identification is classified, and the reduction of energy levels is estimated using an ensemble approach on WSN. The proposed DEEC with Link state-machine learning (LS-ML) approach has improvised the design parameters such as energy, entropy, and packet loss for each data transmission with 8 percent values observed for the improved factors.