Content Distribution Driven By Users' Interests In Opportunistic Mobile Networks

The Internet has developed into an information sharing platform and various web services it provided become enriched gradually. Large amounts of data access lead to considerable web data redundancy and low performance of transmission. Meanwhile, the number of mobile users that participate in the internet access is increasing rapidly. According to statistics, the data traffic consumed by mobile users account for 20% of all the Internet traffic at present, and the proportion is rising continuously. Therefore, how to reduce the traffic load brought by mobile users becomes particularly important.According to the requirement of efficient content distribution in Opportunity Mobile Network, the thesis proposes a content distribution stratery driven by users' interests. To be specific, the thesis combines the content forwarding protocols of mobile opportunistic networks with mechansims of publish/subscribe system. It uses the short-distance wireless capability of mobile devices to achieve efficient and accurate data dissemination. The main work of this thesis is as follows.(1) The design of content distribution process driven by users' interests. According to the content distribution scenario, the thesis proposes content distribution strategy driven by users' interests and analyses data communication process between nodes. It divides the content distribution into four steps incluing the expression of interest and content, the diffusion of interests, the matching algorithm of interest and content, matched content distribution. For each processing step, the thesis discusses the issues to be addressed and candidate approaches.(2) The identification and adaptation of content and interests. The content distribution strategy is determined by the matched results of user interests and content in the system. Different data format correspond to different matching algorithm. The thesis uses information tags to define users' interest format and content structure with corresponding matching algorithm based on Hash encoding. The thesis transfers the interest and content tags into bit vector and then implements the corresponding adaptation algorithm by logical arithmetic of bit vectors.(3) The solution of data interaction between different nodes. The thesis proposes the interest interaction strategy and content interaction strategy by considering the uncertainties of nodes moving characteristic in opportunistic networks. The thesis proposes Known Vector method to implement interest interaction and uses Bloom Filter method to implement content distribution. The strategy prefers transferring interests to transferring content when nodes meet each other. In addition, the thesis defines different data format to transfer data, such as the XML format, the character string and the bit stream.(4) The implementation and measurement of prototype system. The system achieves corresponding functions on fix terminal based on Windows operation and mobile terminal based on Android operation. Different terminals use Bluetooth technology to communicate with each other. In addition, the system has a detailed testing strategy on functionality and performance. The functionality testing contains the test of matching algorithm between interest and content, the test of communication on Bluetooth service. The performance testing focused on content redundancy and communication overhead. The result shows that the system can achieve efficient content distribution in opportunistic mobile networks.