Interest Characteristic Probability Predicted Method in Social Opportunistic Networks

Abstract

In social opportunistic networks, mobile devices can be regarded as socialization nodes. Furthermore, they carry and store useful information. Mobile devices can select destination nodes and deliver messages through social opportunistic networks because messages can be securely and conveniently stored, carried, and transmitted with nodes. According to such characteristics, communities can be established. However, many communities may deliver messages only depending on one or two nodes. If those nodes are not enough cache and over-flooding, data transmission in communities may wait for a long time. To find many cooperation neighbor nodes can share much useful information. This paper proposes a method that characteristic interests with neighbors are selected or pushed service by users. According to establish interest characteristic probability predicted method, interest neighbors can be selected and receive or send information. Compared with the social opportunistic network algorithms, the new research method achieves better results in accuracy selected and delivery ratio, transmission delay, and routing overhead. According to the simulation experiments, the average delivery ratio of new research method is 0.88, which is 40% higher than that of the Epidemic algorithm; the end-to-end delay reduces 38% with Spray and wait for algorithm.