The Research Of Delay Tolerant Network Routing Algorithm In Social-based Network

Paper analyzes the routing problem of social_based Delay Tolerant Networks (DTN). Proposing a social_based DTN routing algorithm based social network features of community,center, similarity and so on.With the mobile phone, PDA and other handheld devices a large number of widespread, using of these handheld devices for data exchange of networks and provides network services which has broad application prospects. Intel Research and Cambridge proposed a PSN (pocket switched network) which is made up of people carry handheld devices formed DTN network. PSN network, however, is not always connected, which makes the traditional routing protocols can not be applied to such network. This is due to the traditional routing protocols, such as AODV, DSR, DSDV and so on, are assumed that the network is fully connected. If the path between the source node and destination node is not existent, information deliver will be failed, which is a big challenge for the PSN network routing algorithm design. To solve this problem, routing algorithm need to adapting the "store - carry - forward" routing mode that transmission message by a hop-by-hop way which formating of connection communication opportunities between mobile nodes and other nodes, to achieve communication purpose among nodes.PSN network is made of people carrying mobile devices, and thus the various features of human society is helpful for the design of social network routing algorithms. Sociological studies have shown that human society is structured by the mutual cooperation among the people of the community which is an important feature of society networks, the idea of correlated interaction means that an organism of a given type is more likely to interact with another organism of the same type than with a randomly chosen member of the population. Community structural change more slowly than social network topology changes, and thus it can be used to do a better forwarding decision. The local community structure can be understood by Community detection algorithm from the off-line analysis and on-line application, and providing great help for good routing design strategy. Sociological research found that: in the community, some people even more popular, and with other people to have more interaction, this level of popularity is also an important feature of human society. It represents the importance of node that as a potential transmission relay in PSN.In this paper, a social-based DTN routing algorithm(SR) is proposed base on the characteristics of human society feature. Social network is divided into different communities by SR algorithm, when message will be forwarded, the community or communities forwarding strategy is chosen by the community which destination node belongs. The degree center, indirect center, similarity, connection forecast and other factors were took into account by forwarding strategy. Dartmouth College data sets were analyzed In this paper, the node type and movement pattern detection by the heterogeneity of social networks and the regularity of people's lives, the routing performance was analysised by community and predicting the possibility of links of nodes by movement patterns. SR algorithm was compared with existing algorithms MEED, Epidemic.Simulation results show: SR algorithm based on social network characteristics in the average delay and the delivery rate aspects has better performance.