User Behavior Prediction Model Based on HMM in HCN

In the context of heterogeneous cellular networks (HCN), to solve the difficult problems that traditional handover management strategy rarely comprehensively considers the mobile preferences and features of users in the hot-spot areas, based on hidden Markov model (HMM), an approach of sensing hot-spot area user behaviors was proposed in this paper. First, based on self-similar least-action human walk (SLAW), the movement paths of hot-spot area users were simulated and a modeling for users was created by using HMM. Then, the movement time was predicted by referring to users' movement sequence. Finally, simulation experiment was conducted to analyze the impact of different sampling time and base station densities on behavior predictions. As a result, specific setting parameters were provided for designing reasonable handover management plans. It turns out that this approach improves the accuracy of the prediction of user movement time to make sure that hot-spot area base stations can be properly prepared for the upcoming user switch request.