The Research Of Energy Efficiency Optimization In Wireless Networks Based On Geographical Location Prediction

With the popularity of mobile terminals, the location based social network appeared in our life. Compared with traditional social network,the location based social network provides users with related location-based service, it allows users to get their current location via their mobile devices and take part in social activities with these location. At the same time, with the increasing development of telecom network scale, mobile communications technology's energy consumption is increasing. How to reduce the wireless network's energy consumption and optimizing the wireless networks's energy efficiency are the main goal of the current information and communication technology. However, the traffic of the wireless networks is uneven in time and space, and the deep reason is that user mobile behavior is uneven in time and space. Considering the relationship between social network and wireless networks, this paper dealt with the location based social network, and focused on the problem of energy efficiency optimization in wireless networks.The location based social network contains a large amount data of user's geographical location. By analysing these data from the location based social network, we can abstract underlying rules in user mobile behavior and service demands. Furthermore, it offers a new reference dimension for energy efficiency optimization in wireless networks and achieves proactive management of radio resource. The contributions of this paper included the following two aspects.First, analysis of characteristics of user's geographical location data and research on the prediction algorithm. As to analyzing the data, this paper made research on it from three aspects, which is temporal, spatial and cyclical. Some rules had been found and had been used to predict user's location. And in order to solve the problem of data sparsity in this research which makes low accuracy, this paper proposed a smooth algorithm which is based on temporal and cyclical rules. The algorithm adopts the least square method for curve fitting and experiments proved that the algorithm had improved accuracy.Second, this paper proposed a high energy efficiency wireless networks services which is based on location prediction. This paper proposed a location-traffic correspondence model which is combined with user mobile behavior and location prediction. By controlling the base station's open or sleep, it can reduce the energy consumption of wireless networks. Meanwhile, this paper built up an energy efficiency optimization scene based on location prediction, which had been proved by experiment that our model is feasible. The experimental results showed that the algorithm proposed by this paper can reduce the energy consumption in wireless networks.