基于MDP的Massive MIMO物理层安全算法

MASSIVE MIMO PHYSICAL LAYER SECURITY ALGORITHM BASED ON MDP

对Massive MIMO(multiple input multiple output)系统的物理层安全问题进行研究,提出基于马尔科夫决策过程MDP(Markov decision process)的物理层安全模型。利用动态规划DP(dynamic programming)求解MDP模型。利用互阻抗模型建立Massive MIMO系统下行链路的信道模型;建立基于平均无折扣回报的MDP模型,利用有限状态马氏信道FSMC(finite state Markov channel)的区间转移概率模型,给出信道的转移概率表达式;提出基于值迭代算法的MDP物理层安全算法,得到加密容量最大化的全局最优策略;通过仿真对算法性能进行验证。仿真结果给出基站发射信号功率对信道物理层加密容量的影响,并评估算法的性能。

We studied the physical layer security of massive MIMO(multiple input multiple output)system,and proposed the physical layer security model based on MDP(Markov decision process).The MDP model was solved by DP(dynamic programming).We established the channel downlink model of massive MIMO system by the mutual impedance model,and set up the MDP model based on average no-reward returns.The transfer probability expression of channel was given according to the interval transfer probability model of FSMC(finite state Markov channel).We also presented a MDP physical layer security algorithm based on value iteration to obtain the global optimal strategy with maximum encryption capacity.The performance of the algorithm was verified by simulation.The simulation results show the influence of the base station transmitting signal power on the channel physical layer encryption capacity,and evaluate the performance of the algorithm.