Jarvis-Patrick-Clusterative African Buffalo Optimized Deep Learning Classifier for Device-to-Device Communication in 5G Networks

Abstract

ABSTRACT With the fifth generation (5G), there are several mobiles devices may connect to the network from any location. Device-to-device (D2D) communication is a crucial component of 5G wireless networks since it increases performance. D2D communication in the 5G networks is the subject of extensive study. However, the existing D2D communication techniques did not show an improvement in energy economy or a reduction in latency. The Jarvis-Patrick-Clusterative African Buffalo Optimised Deep Learning (JPCABODL) Model was created to overcome the problems. The JPCABODL Model's main goal is to improve D2D communication performance in 5G networks using a deep learning classifier with improved energy efficiency and lower latency. Four layers make up the JPCABODL Model: one output layer, two hidden levels, and one layer for device-to-device communication. A layer of input is the total number of mobile devices. After that, each mobile device's received signal strength (RSS) and residual energy are taken into account when the Jarvis-Patrick-Clustering procedure is carried out at hidden layer 1. After that, each cluster's cluster head is selected from mobile devices with the highest residual energy, bandwidth, and RSS. The cluster is then transmitted to hidden layer 2 together with information about the cluster leader. African Buffalo Optimisation is used in that layer to choose the ideal cluster head. In 5G networks, effective device-to-device communication is accomplished with the aid of an ideal cluster head. The JPCABODL Model and current techniques are used for simulation, and several metrics such as energy efficiency, data delivery ratio, packet loss rate, throughput, and latency are measured. When compared to traditional approaches, the experimental evaluation of the JPCABODL Model enhances EE, DDR, and minimizes latency.