A reinforcement learning-based intelligent route selection scheme is proposed to address communication security issues in randomly distributed multi-hop relay networks.The scheme considers respectively single-key and independent-key transmission scenarios and investigates the detection performance at the wardens,and the optimal power allocation for each relay is determined based on covert constraints.Finally,an intelligent route selection scheme for multi-hop network transmission is implemented based on reinforcement learning techniques to ensure the covert transmission and maximize the system's covert throughput.The results show that the route selected by the single-key scheme tends to avoid the area being monitored by the wardens,while the route selected by the independent-key scheme can pass through the wardens'monitoring area.Moreover,the system gain achieved by the independent-key scheme is significantly better than that of the single-key scheme.
维普
万方数据