A communication security anti-interference decision model using deep learning in intelligent industrial IoT environment

To traditional anti-jamming decision algorithm that cannot meet the security needs of smart city development, this paper proposes a communication security anti-interference decision algorithm using deep learning in an intelligent industrial IoT environment. Firstly, an interactive system model of cognitive users and disruptors with intelligent perception function is constructed. Besides, the interference intensity and channel gain are comprehensively analyzed to design the optimization goal to maximize network capacity. Then, by modeling the interaction between cognitive environment and decision engine as the interaction between environment and agent in deep reinforcement learning, the Q-learning algorithm integrating reinforcement learning is used to explore the maximum action reward feedback to cognitive decision engine, so as to intelligently obtain the effective interference parameters of communication state. Finally, the proposed algorithm is experimentally demonstrated based on MATLAB simulation platform. The results show that when the number of links is 300, the network capacity of proposed algorithm is about 960 \(\text{bit} \cdot \text{s}^{ - 1} \cdot \text{Hz}^{ - 1}\), and the cumulative average reward value reaches 0.59, which is better than the comparison algorithm, and realizes high reliable autonomous decision-making.