An efficient self-attention-based conditional variational auto-encoder generative adversarial networks based multipath cross-layer design routing paradigm for MANET

Abstract

Battery life is often a helpful resource for portable devices in MANET (Mobile Ad Hoc Network). When creating control methods, energy efficiency with network lifespan is considered. However, creating a routing method that is energy-efficient requires taking into account a variety of nodes from various layers, including channel assumptions, total traffic load, and remaining energy. The traditional layered technique cannot tackle power-relevant problems that may affect every stack layers. In this manuscript, Self-Attention-Based Conditional Variational Auto-Encoder Generative Adversarial Networks is proposed for enabling cross-layer routing over Mobile Ad Hoc Network (SACVAEGAN-MCLD-MANET). This Self-Attention-Based Conditional Variational Auto-Encoder Generative Adversarial Networks (SACVAEGAN) choose the optimum path by deeming given objective functions: residual energy cost of communication, Data Success Rate (DSR), mobility. In which, energy, DSR, mobility are computed through network layer. MAC layer's hesitating fuzzy linguistic bi-objective clustering technique (HFLBC) reduces the energy consumption brought on by contentions. The nodes with lesser residual energy must complete the immediate data transfer, while the path with greater contention must wait. As a result, this HFLBC is used to reduce energy use and enhance data transmission. Adaptability is produced using the knowledge that is shared among the different layers. The proposed method is done in Network Simulator-2.34, its efficacy is analyzed under certain performance metrics, like energy consumption, Packet Delivery Ratio, end-end delay, routing overhead, network lifetime, PLR, throughput. The proposed approach provides 23.21%, 15.64%, 19.96% higher packet delivery ratio and 16.63%, 17.29%, 6.47% lower delay compared with existing methods, like Sustainable Multiple path Routing for Enhancing Cross-Layer Presentation utilizing Energy Centric Tunicate Swarm Approach in MANET (ECTSA-MCLD-MANET), Cross-Layer adaptive data scheduling policy in mobile ad-hoc networks (ADSP-MCLD-MANET) and Cross-Layer Design Based N-ary Huffman Coding for DSDV Routing Protocol Analysis in mobile ad-hoc networks (NAHC-MCLD-MANET) respectively.