Based On The Context Of Delay Tolerant Network Routing Algorithm

Typical Delay Tolerant Networks (DTNs) often suffers from the long and variable delays, frequent connectivity disruptions and high bit error rate. In DTNs,the design of an efficient routing algorithm is one of the key issues. Existing routing algorithms can be classified into two categories: replication-based routing and forwarding-based routing. Through sending multiple copies to the network, the former algorithm improves the data delivery, but may cause a high network overhead. The later improves success delivery rate using more network knowledge, but getting all oracle knowledge in network is impossible, because of the frequent changes of network topology. Considering above, we proposed location prediction using Markov prediction model, and studied the routing algorithm about these locations.The contributions of this thesis are as follows:Firstly, using the Markov forecasting mechanism, the thesis proposed the 2-order Markov prediction model. Envisaging the network formed by a number of position points in model, due to the continuity of motion, node can predict the next position, according to its context, the present and previous positions. Transition probability is calculated using the history value recorded by the node. At the same time, the model set the sampling interval, which value is based on network's size and speed of the nodes, can be well-suited to a variety of the large delay network environment.Secondly, proposed the DTN routing algorithm based on Markov position prediction model, called Spray and Forward. Node predicts the destination node's location, and uses a hybrid approach to forward packets. In the routing initial stage, Spray and Forward enters the spray phase of the Spray and Wait, which spray the copies of packet randomly to other nodes; when the node has only one copy of the packet, the node forwards this copy using greedy forward. The introduction of forwarding strategy reduces the number of copies in the network, decreases the network overhead; multi-path forwarding of the packets improves the success delivery rate.Thirdly, proposed the advanced Spray and Forward routing algorithm based on cross-layer. Most routing algorithms do not take into account the limited energy and cache of node, the fourth chapter introduces the advanced Spray and Forward algorithm. The algorithm is that the node routes the packets considering not only the location of neighbor nodes, but also nodes'remaining energy and cache. It can effectively improve the network lifetime and success delivery rate, and reduce network overhead.And the thesis evaluated the performance of the advanced Spray & Forward and Spray & Forward using OPNET simulator. Finally, concludes the thesis and indicates the future works.