Research on Channel State Prediction Algorithm under Multi-radio Multi-channel Environment in Heterogeneous Networks
异构网络中多无线电多信道环境下信道状态预测算法研究

Due to heterogeneous network environment with the characteristics of uncertainty for unlicensed users, based on the theory of Partially Observable Markov Decision Process (POMDP) model, a novel multi-radio multi-channel channel state prediction algorithm is proposed using spectrum sensing techniques of cognitive radio. By the analysis of the channel state historical information, the initial distribution of the belief state and transition probability is derivated and the channel with optimal reward is selected for unlicensed user to improve the spectrum utilization. The simulation results demonstrate that the proposed algorithm has better performance than classical algorithms.