Research On The Key Technology Of Networks Resource Management Based On Human Mobility

With the rapid development of wireless network technology bringing more convenient services to people's life, the number of intelligent mobile terminal users increased sharply so that a variety of mobile applications and services come into people's life and diversified information can spread in the network rapidly. Therefore, how to guide the efficient spread of information between terminals and improve utilization rate of terminal resources have drawn much attention of people.Meanwhile, the explosive growth of wireless network data traffic resulted in the increase of network overhead and the drastic decrease of network performance. In order to meet the need that more users can access the network and ensure the service quality of the network, people start to pay attention to how to reduce network load and improve utilization rate of network resources by offloading wireless network data. At the same time,people begin to deploy network resources on a large scale with the aim of improving network capacity. However, the deployment of various network resources caused a lot of energy consumption which means huge operation cost. Hence, designing an efficient dynamic resource management mechanism to reduce energy consumption has become the hot topics of current research.As the carrier of intelligent terminal, humans' movement will cause the changes of physical location which will lead to the contact of people and thus huge information exchange. At the same time, the periodicity and tidal characteristics of human mobility will lead to the uneven dynamic distribution of the traffic load of wireless network to a certain extent. Based on characteristics of human mobility, this paper first studies the influence of characteristics of human mobility on information spreading between mobile terminals. Then with the use of contact characteristics of human mobility, this paper designs a network traffic data offloading mechanism by means of opportunistic communication.Finally, by analyzing spatial and temporal characteristics of human mobility and network traffic load, the energy efficient dynamic resource management mechanism is designed. The main innovations and contributions of this paper are as follows:(1) Measuring the spread influence of nodes based on characteristics of human mobility. This paper uses the way of direct communication between terminals. In order to ensure the rapid transmission of information between terminals and a large transmission area, it is necessary to select a mobile terminal with a larger spread influence to distribute messages. Different from existing work, this paper propose measures spread influence of a node based on step size and radius of gyration of human mobility. By using a large scale real world human trajectory data and with the use of SIR (Susceptible-Infected-Recovered)we identify the ability of information spread. Simulation results show that information spread influence doesn't increase with the increase of spreader's step size and radius of gyration, instead, when step size and radius of gyration achieve median value we can get the greatest spread influence. Hence, according to the range of spreader's step size and radius of gyration we choose the target node to achieve efficient information spread and hence improve the utilization rate of terminal resources.(2) Designing network data offloading mechanism based on characteristics of human mobility. The physical location's change due to human mobility will lead to the contact of people and thus huge information exchange. For data that can tolerate large delay, we can offload traffic data with the use of contact characteristics through the way of opportunistic communication. The basic idea is to deliver messages to the target node and then allow the target node spread information to other users through opportunistic communication. Therefore, based on characteristics of contact characteristics, we propose NodeRank algorithm to select the target node. Different from existing researches, we not only consider contact frequency between users, but also combine contact times and inter-contact times to ensure the integrity and availability of message delivery. Based on real-world mobility traces, simulation results show that compared with the traditional method without offloading data our method offloads almost 75% data traffic in cellular network. The efficiency has risen 6% and the large amount of network redundant copies have been reduced compared with the offloading method which only considers characteristics of contact frequency.(3) Designing energy efficient dynamic resource management mechanism based on characteristics of human mobility. As the main body of generating traffic data, the periodicity and tidal characteristics of human mobility in both time and space domain will result in non-uniform dynamic distribution of the wireless network traffic load to a certain extent, which will lead to low load or idle operation of network resources in some region. Hence, based on characteristics of human mobility and fluctuation of traffic data, we analyze the potential causes of energy consumption in wireless networks. From the perspective of reducing the network energy consumption and improving utilization rate of resources,we propose three cell zooming models and three kinds of cell cooperation mechanisms correspondingly. It is confirmed by the experiment that in the situation where the path loss exponent is normal and low outage probability is guaranteed, compared with the traditional model where network resources are equipped with peak value of base station, our method reduces the energy consumption of more than 50%.On the basis of characteristics of human mobility, this paper measures the spread impact of node, designs a network traffic data offloading mechanism by means of opportunistic communication and designs energy efficient dynamic resource management mechanism based on changes of traffic data.