This paper proposes a reinforcement learning-based precoding scheme with artificial noise to enhance secure communication in multi-input multi-output (MIMO) interference channel networks. The system consists of K transmitter-receiver pairs communicating while exposed to a multi-antenna eavesdropper under channel uncertainty. To address the secrecy rate maximization problem, which involves highly non-convex optimization due to power constraints and coupled variables, the problem is formulated as a Markov decision process (MDP) and solved using the deep deterministic policy gradient (DDPG) algorithm. Numerical results show that the proposed approach achieves comparable secrecy performance to the latest asynchronous distributed pricing-based scheme while significantly reducing the computational complexity.

Keywords