The Study Of Nodes Behavior And Multicast In Opportunistic Networks

The concept of opportunistic networks is based on the delay tolerance of network andmobile ad hoc network put forward that one does not need the complete path between thesource node and the destination node, depending on the nodes encounter opportunities,intermittent network communication. The idea of “store-carry-forward” is more suitable forthe actual needs of Self-organizing network routing. Because of the discontinuity of theopportunistic network connectivity, resulting in traditional routing protocols ofSelf-organizing network can not be effectively in the network. Therefore, efficient routingalgorithm with opportunistic network become the focus of people’s research. This paper putsforward two kinds of opportunistic network routing algorithm based on the analysis andsummary of the existing routing algorithms for opportunistic networks, the main researchcontents are as follows:(1)According to the existing opportunistic networks’ routing algorithms, which has theproblems of bindly forwarding and network load, puts forward a kind of network routingalgorithm based on node behavior opportunity. Uses the behavioral information of the nodesto make predictions about their movements in the network. This helps to fnd and select abetter next hop for the message to be routed to the destination. It also incorporates a methodfor the acknowledgment of the received messages, which helps in the buffer management ofthe intermediate nodes.(2)In order to improve the algorithm of describe above in the following work, this articlehas conducted the research to the multicast network based on coding of opportunisticnetworks, put forward a distributed method to obtain the best solution of multicast. Thedistributed solution consists of two phases. In the first phase, the most reliable broadcastingtree is formed based on the ETX (Expected Transmission Count) metric. In the second phase, a credit assignment algorithm is run at each node to determine the number of coded packetsthat the node has to send. The distributed algorithm adapts to the changes in the channelconditions and does not require explicit knowledge of the properties of the network to reducethe number of feedback messages, and to resolve the problem of delayed feedback, we alsoperform network coding on the feedback messages.The above two algorithms respectively from the context information and the node innetwork coding multicast makes the improvement to the existing algorithm, simulation resultsshow that these two algorithms have achieved excellent performance in terms of networkperformance.