Social-based DTN Routing Mechanism And Application Of Network Coding

The concept of network coding was proposed in 2000, and the multicast max throughput can be obtained through network coding. Network coding adopts a new strategy of store-code-forward mode instead of the traditional store-forward mode. Network coding can improve throughput, load balancing, robustness, security and reduce the delay. Network coding can also improve the reliability of wireless networks. Therefore, network coding has attracted lots of research interests and has been applied widely in wireless networks.In this thesis, we focus on social-based DTN routing mechanism and application of Network Coding. CS-DTN is an abbreviation of network coding based clustering and social network protocol in Delay Tolerant Networks. CS-DTN performs well while transmitting data between nodes which are placed and move based on social nature in DTN networks. CS-DTN clusters nodes using of the encounter probability and identifies centricity using node activity. There are different routing mechanisms between intra-cluster and cluster-to-cluster. Furthermore, CS-DTN combines network coding and data generation & forward together to improve the performance. This paper adopts random linear coding scheme, and combines source node coding and hubs coding which have higher rank of centricity. This method can reduce the coding pressure on source node, improve the effect of random coding and reduce redundant data.At last, CS-DTN is implemented on QualNet simulation platform and compared with Clustering protocol, Epidemic which is a traditional DTN routing protocol and HubCode protocol. The results show that CS-DTN outperforms Clustering in terms of delivery ratio and average delay. Network coding for CS-DTN have a good optimization results, which can improve the delivery ratio of nearly 10%. In addition, to achieve the best results, top 20% of the nodes which ranked by centricity should be set as Hub nodes.