Intelligent QoS-Driven Ad Hoc On-Demand Distance Vector Routing for 5G MANET Using Physically Recurrent Neural Network

Mobile ad hoc networks (MANET) continue to evolve within the 5G era, which has gained increased attention. Its primary characteristic is that nodes are constantly applied to heavy traffic loads and the QoS requirements are necessary. The routing protocols often struggle to maintain high QoS under dynamic and unpredictable network conditions. It is difficult to create a routing protocol that effectively adjust to node mobility, changing traffic and fluctuating network quality while maintaining crucial QoS metrics. To address this challenge, this manuscript proposes an intelligent QoS-driven AODV routing for 5G-MANET using physically recurrent neural network (AODV-5G-MANET-PRNN). The PRNN is used to update the traffic loads on database and to improve QoS-ensured route destinations. Then, the database fed to AODV-PRNN detects the guaranteed QoS in routing. The proposed AODV-5G-MANET-PRNN technique is implemented and analyzed using performance metrics like end-to-end delay, throughput, network lifetime, energy consumption, packet delivery ratio (PDR), signal-to-noise ratio (SNR), jitter, delay variance, and computational cost. The proposed approach attains 19.68%, 22.34%, and 30.22% higher throughput; 9.75%, 14.86%, and 10.42% lower Jitter; and 9.44%, 12.38%, and 7.29% lower delay variance compared with existing AODV-EQOS-5G-MANET, RP-QOS-DDE-5G, and QOS-5G-EEC models, respectively.