Research On Data Delivery And Resource Management In Mobile Social Networks

Mobile Social Network (MSN) is a distributed mobile wireless network composed by users of mobile devices within a particular area. For instance, the students of a campus can compose an MSN with their mobile devices. In MSN, adjacent users communicate with each other using short-range communication techniques, such as WiFi-direct and D2D, without connecting to infrastructure networks, such as cellular network. Nonadjactent users exchange data through the forwarding of other users. With the aid of MSN, users can disseminate and share information directly. This can be used to support some particular services, such as advertisement spreading. Besides, it can also offload the traffic of infrastructure networks and improve the capacity of the infrastructure networks.In MSN, data delivery is the most important issue. However, due to the mobility of users, data delivery is also a big challenge. Traditional routing based data delivery techniques can no longer be used in MSN (since no fixed route can be established). Hence, new data delivery techniques with high efficiency need to be explored. On the other hand, as a mobile wireless network, network resource (device resource, channel resource, etc.) management is another critical issue. A highly-efficient resource management technique can not only save network resource and reduce the cost of communications, but also improve the efficiency of resource and enhance the performance of the network.Therefore, this dissertation focuses on data delivery and resource management of MSN. Two data delivery methods and three resource management approaches are proposed. The main contribution of this dissertation can be summarized as follows.(1) Based on the social strength among users, a new protocol (Social Strength based Dissemination protocol, SSD) is proposed for dissemination and sharing of multiple interest data. Firstly, a data structure is designed, based on which a data collection method is proposed to collect the encounter patterns among users. Then, a metric is developed to evaluate the capability of a user to be the relay of particular interest data. Finally, the SSD protocol is proposeda. It is validated by simulations that, SSD can discover the optimal relay of data based on the social strength among users, and thus improve data delivery efficiency.(2) Further considering the social strength between users and locations, a new protocol (Integrated Social Strength based Dissemination protocol, ISSD) is proposed for data dissemination and sharing. Firstly, the social strength among users and the one between users and locations are computed respectively, based on which the integrated social strength of users is derived. Then, based on the integrated social strength of users, ISSD protocol is proposed. In order to analyze the properties of ISSD protocol, a continuous-time Markov process based analytical model is developed, which proves that ISSD has a small cost and a good scalability. It is validated by simulations that, ISSD protocol has a big improvement in the performance of delivery ratio and delivery delay.(3) Storage management of stationary relay devices is explored to improve the performance of data storing and forwarding on relay nodes. According to the deployment of relay devices, the storeage management problem is classified into two more specifical problems:storeage allocation of stationary relay devices and storage allocation of deployable relay devices. In order to solve these two problems, social strengths between users and potential locations of relay devices are evaluated based on the visit pattern of users. Then, the demand of storage of each location is estimated according to the data storing history of each location. Finally, two optimization models are proposed to calculate the two problems, respectively. It is validated by simulations that, the two storage allocation scheme can allocate storeage to each relay device efficiently and thus improve the performance of data delivery.(4) Energy management issue is studied in combination with wireless energy harvesting techniques. By applying wireless energy harvesting techniques into stationary relay device, a new relay device called E-Box is proposed, which can simultaneously transmit data and energy. By deploying E-Box, data delivery efficiency and energy harvesting efficiency can be optimized. First, based on the analysis of user mobility, user mobility model is established. Then, the average sojourn time of each user at each potential deployment location is derived using Markov stable state analysis. Finally, based on the average sojourn time, three E-Box deployment schemes (D-deployment, E-deployment and T-deployment) are proposed for maximizing data delivery efficiency, maximizing energy harvesting efficiency and making a tradeoff between them, respectively. It is validated by simulations that, the three deployment schemes can efectively realize their goals and improve the efficiency of data delivery and energe harvesting.(5) A Coalition based Bandwidth Allocation (CBA) technique is proposed for the base station to allocation bandwidth to the users in an MSN. Firstly, based on the analysis of user mobility, user mobility model is established. Then, a profit function that considers the throughput and delay of users, is proposed to evaluate the profit of users in coalition formation. Finally, based on the profit of each user, CBA is proposed for coalition formation and bandwidth allocation. It is validated by simulations that, with rational coalition formation, CBA can largely improve the throughput of users.