Analysis And Prediction Method For Wireless Communication Networks Based On Space-time Demension

With the rapid development of Information Communication Technology, network applications and scales expand increasingly. The number of base stations increased rapidly. At the same time, the difficulty of base station management as well as the costs of wireless network maintenance is secular growth. That improving the resource scheduling efficiency of the base station and the efficiency of wireless access and resource utilization is of great significance for the sustainable development of the industry. Since the optimization of the base station configuration depends on the characteristics of the base station traffic, the research and prediction of base station service quantity has become a very important part of the optimizing the configuration of the base station.With the rapid development of data mining and other methods in recent years, these analysis theory has not yet been applied to the analysis of the traffic analysis of the base station. At the same time, although the existing traffic forecasting methods are based on the change of the current network, they also have some deviations and limitations.Based on the above problems, first we analyses the base station traffic using statistical and cluster analysis based on the two dimensions of time and space. Based on the actual traffic data of base station, we give a detailed statistical analysis of the distribution of base station traffic in two dimensions of time and space. On this basis, a method of base station aggregation based on traffic and geographical position is proposed using the theory of data mining and the traffic characteristics and geographic location information of base station.Secondly, we proposes a more accurate prediction method for the voice traffic and the data traffic. Aiming at the characteristics of voice traffic, a mathematical model based on time series analysis is used in this paper and the predictive accuracy is improved compared to other forecasting methods currently in use. Based on the time series decomposition method, we prediction the data characteristics of data traffic and obtains good results.In conclusion, based on station traffic analysis research and traffic forecast,a better base station aggregation method and traffic forecasting model are proposed that can provide better support for the work of energy saving management and base station dispatch of base station so as to achieve intelligent regulation, reduce the cost of human intervention purposes.