Security Enhancement and Minimal Usage of Energy in MANETS
Abstract
The Despite Mobile Ad Hoc Networks (MANETs) being an important communication paradigm, the intrinsic properties of MANETs, such as dynamic nature, lack of sufficient energy, and susceptibility to attacks, make it difficult to ensure security as well as improve efficiency. In this paper, a new framework for the enhancement of security as well as energy efficiency is presented by combining Quantum Graph Neural Networks (QGNN), Kolmogorov-Arnold Networks (KAN), and Twin Delayed Deep Deterministic Policy Gradient (TD3). The system is simulated using NS-3 in order to integrate different routing protocols (AODV, DSR, OLSR) along with energy and mobility models. Attacks are simulated in order to measure security and then analyze the logs with the help of Python. The approach of applying QGNN to learn the structure of the communication graph, KAN to approximate the function related to the complex traffic patterns, and TD3 to refine the policy is shown to be effective in detecting anomalies in a timely fashion without wasting any energy unnecessarily. The experimental results demonstrate that the proposed system can deliver an accuracy rate of 97.82%. In general, this paper introduces a mixed intelligence-driven technique that fulfills the requirements of securing MANETs and extending their lifetime. Through the amalgamation of graph intelligence in terms of structure and adaptive reinforcement, the proposed design manages to find an effective balance between both security and efficiency. The simulation results emphasize the practicality of the approach by ensuring reliable performance regardless of the routing and mobility scenarios. In conclusion, this work shows how hybrid intelligence can be used for securing communications in decentralized wireless networks.