Optimizing Mobile Ad Hoc Network cluster based routing: Energy prediction via improved deep learning technique

Abstract

SummaryMobile Ad Hoc Network (MANETs) are indeed autonomous, fast‐deployable wireless networks that are ideal for communications in areas with limited radio infrastructure, outdoor events, military operations, and disaster relief. But the primary problem is with routing because of the nodes' mobility. This paper proposes a cluster‐based routing under energy prediction via deep learning techniques. Here, proposes a new model termed as Concatenation of Convolutional with Max‐Avg Pooling layer in Deep Convolutional Neural Network (CCMAP‐DCNN) for energy prediction, in which additional layers are inserted into the extant DCNN structure to ensure the effectiveness of energy prediction. As a result, clusters are created by grouping the nodes together. Constraints like energy, trust, distance, and latency are taken into account while choosing the cluster leader. For cluster head selection and optimal routing, proposes a new hybrid Namib Beetle Upgraded Jellyfish Search Optimization (NBUJSO) algorithm that utilizes the NBO strategy to the JSO algorithm making the selection process more optimal. The best route is then chosen by the optimum routing method, which takes into account variables like network quality and mobility to send data packets from source to destination as efficiently as possible. Finally, the data aggregation process is followed for eliminating the redundant transmission.