Machine Learning-Driven Rate Adaptation Strategies for Cross-Layer Routing in MANETs
K. Sathish, D. Chitra, Rajesh Sharma R, Akey Sungheetha, L. Rajavimalanaathan, Vijayan Ellappan
Abstract
A Machine Learning-Driven Rate Adaptation Strategy for Cross-Layer Routing in Mobile Ad-Hoc Networks (MANET) represents the main contribution of the research which solves QoS challenges in real-time applications. The protocol uses machine learning algorithms to detect dynamic changes in network conditions which include link quality assessment and queue status monitoring and congestion monitoring. This network design unites data from physical along with MAC and network layer components to reach optimal resources utilization. Real-time traffic patterns together with node mobility data get processed by machine learning models which lead to adaptive congestion control and rate adaptation decisions. Through its approach the protocol identifies several node-disjoint paths which boost network resistance to link failures. Computer simulation results prove that the protocol delivers superior packet delivery performance and lower end-to-end delay and enhanced adaptability under changing traffic conditions and mobile node environments. The new methodology surpasses conventional methods by offering strong backing for applications that require delayed tolerance and vast bandwidth needs while handling modern MANETs so well through a complete performance solution.