Slime Mould-Based Collaborative Deep Boltzmann Machine for Intrusion Detection Model in Mobile Ad Hoc Network

Abstract

Mobile Ad Hoc Network (MANET) is a dynamic wireless network developed by using wireless nodes without using any infrastructures. The significant features of MANET are low-cost infrastructure, self-organization, mobility and rapid deployment, which offer the opportunity to deploy it in various applications such as disaster relief, environmental monitoring and military communications. Based on the previous studies, improved Quality of Service (QoS) metrics with security problems during data communication are challenging with the increased wireless technology. By addressing these issues, here proposed a novel secured intrusion detection model in MANET. The cluster formation is effectuated with the Modified K-Harmonics Mean Clustering (MKHMC), and the cluster heads are selected with the proposed Chaotic Multi-verse Krill Herd Optimization (CMKHO) algorithm, which helps to provide energy efficiency, reduction in delay, and increased throughput. Meanwhile, this proposed blockchain-secured Slime Mould-based Collaborative Deep Boltzmann Machine (SM-CDBM) includes three stages, (i) learning the unimodal DBM models to identify the intrusion, (ii) learning the shared layer parameters utilizing a Collaborative Restricted Boltzmann Machine (CRBM), and (iii) fine-tuning the CDBM using the Slime Mould Optimization (SMO) algorithm. Simulations are effectuated in the NS2 tool and accomplish improved malicious node detection, end-to-end delay, energy efficiency, and overhead compared to other state-of-the-art approaches.