Q-Learning Based MEP Search Algorithm and Coverage Enhancement Strategy in IoT-Enabled Intrusion Detection

The minimum exposure path (MEP) represents the worst case of the whole network coverage, how to find MEP adaptively in the network environment with the lowest network overhead and improve the network intrusion detection ability based on MEP is a challenging task. In this context, firstly, we propose a minimum exposure path search algorithm based on the adaptive Q-Learning (MEP-AQL) to solve the path finding problem in the directional sensor networks. By constructing a weighted grid model, the traditional MEP problem is transformed into a path planning problem, and aiming at the Q-Learning (QL) algorithm is difficult to balance exploration and utilization in action selection, its action selection strategy is improved. Secondly, in the coverage enhancement stage, a MEP-based assignment coverage enhancement strategy is proposed, which realize network coverage enhancement by minimizing the mobile energy consumption during the scheduling process of security redundant nodes in the network. The simulation results show that the MEP found by MEP-AQL algorithm has better performance than the QL algorithm and ant colony algorithm. After using the coverage enhancement strategy, the network coverage rate is increased by 35.13% and the energy consumption of node movement is reduced by 47.58%. Therefore, the strategy proposed in this paper can effectively improve the monitoring ability and coverage performance of the network at a lower cost.