Poster: A combined prediction handover algorithm with mobility history in Heterogeneous networks

Prediction handover based on user mobility is one of the most efficient strategies to reduce the handover error ratio and future implement the fast and seamless handover in Heterogeneous networks (HetNet). In this method, UE's history movement measurements were decomposed into trend part and non-trend part, the two parts were predicted by AR model and fuzzy neural network (FNN) respectively to improve the prediction accuracy.