PREDICTIVE INTELLIGENCE FOR RESILIENT HIGH-MOBILITY MANET TOPOLOGIES

The frequent changes in the topology, dynamic movement of nodes, and intermittent connectivity along with unpredictable link failures in High-mobility Mobile ad Hoc network (MANET) have great impact on the reliability and performance of the routing. Reactive (usually conventional) routing protocols trigger high route discovery overhead, energy consumption, latency or packet loss during topology changes. In this study, the authors introduce an idea based on the use of the so-called Predictive Intelligence in MANET topologies with high mobility to complement the mobility prediction with a link-quality estimation and with an intelligent selection of the routing path, which anticipates that a certain communication route may break down. This framework consists of analyzing mobility patterns at each node, stability of the links, available residual energy at each node, congestion, and neighborhood characteristics to predict the potentiality of each link disconnection and dynamically constructing a stable and robust route. Predictive analytics using machine learning helps to detect mobility- and link-failure patterns, while an adaptive routing mechanism favors stable and energy efficient paths. Evaluation results of the proposed approach are compared with packet delivery ratio, end-to-end delay, routing overhead, throughput, energy consumption, and route lifetime of conventional MANET routing protocols based on different mobility conditions. The framework will be expected to give stronger resilience to every routing operation in addition to minimizing route reconstruction and communication loss in highly dynamic network environments including different fields such as vehicular networks, wireless networks, and networks based on mobile agents. The proposed predictive approach is intelligent, and serves as a solid basis for MANET communication self-adaptivity and resilience especially in application scenarios such as disaster response, vehicular communication, military communication, and other infrastructure-independent cases.