Research Of Mobility Prediction Algorithm For Optimized Resource Allocation In Heterogeneous Networks

In recent years,with the development of mobile communication technologies and the popularity of smart phones,there are rocket-increasing mobile subscriptions.At the same time,the user's demands and the mobile data traffic grows explosively,,which lead to higher requirement for the mobile network capacity.The deployment of small base stations in hot areas and the edge of macrocells can improve the network capacity by forming the heterogeneous network with macro base stations.However,due to the mobility of users,the user equipment would handoff frequently.The efficient user mobility prediction algorithm can help optimize the handoff,resource allocation and network load balance,which is one of the hot research area.This paper research on the user mobility prediction algorithm in the heterogeneous network,and the main contents are as follows:This paper proposes a mobility prediction scheme based on order-2 Hidden Markov Model(HMM)to optimize the resource allocation.There are some deficiency in existing HMM prediction model with Viterbi algorithm.This paper proposes an essential prediction unit to execute the prediction process for improving the accuracy and efficiency of prediction.We propose the mobility prediction scheme based on order-2 HMM,which considers not only the impact of previous states on the next location prediction,but also the corresponding relationship between hidden and observable states,which would be beneficial for enhancing the prediction accuracy.Both the time and spatial factors are considered in order-2 HMM to improve the prediction results.The user's service requirement in next location is analyzed and the network allocates resource appropriately for the user.Finally,the simulation results show that the proposed scheme achieves good performance in enhancing mobility prediction accuracy,improving resource utilization while meeting user's requirement.This paper proposes a resource allocation scheme based on mobility prediction to improve the network load balance.Based on the analysis of user's movement history,we establish the mobility prediction model with decision tree.To find the beneficial input training attributes,we study the effect of user's previous several movement states on the prediction to improve the prediction accuracy.At the same time,we consider not only the space change,but also the time factor to enhance the accuracy.Then,the time series analysis method is adopted to find the load status of base stations when the user arrives at the next location,and we use the Autoregressive Integrated Moving Average Model(ARIMA)to analyze the cell's load status.We present the resource allocation scheme to optimize the network load balance based on the user mobility prediction result and target cell load status analysis.Eventually,simulation results show that the proposed scheme could improve the mobility prediction accuracy and optimize the network's load balance effectively in the heterogeneous networks.