Research On Routing Algorithm In Opportunistic Networks

Opportunistic Networks is a novel self-organizing network model with the general characteristics of the Delay Tolerant Network and it utilizes the encounter opportunities arising from node movement to implement communications and aims at resolving the data communication problem for the frequent interruption network. In Opportunistic Networks, nodes usually form a disconnected network due to nodal mobility, low density, limited energy and storage, etc. The source node and destination node being located in different connectivity domains leads to the result that conventional communication protocol of wireless ad hoc networks can not be effectively run in Opportunistic Networks.To solve the above problems, this thesis addresses on routing algorithm for Opportunistic Networks according to its characteristics. To begin with, the technical background and some current typical applications of opportunistic networks are introduced, which follows the elaborating of some routing algorithms and the analysis mainly on the principle and performance of PROPHET routing algorithm based on a brief analysis of the advantages and disadvantages of various routing algorithms. PROPHET algorithm, based on epidemic forwarding and encounter-based forwarding routing mechanism, indicates that each node estimates forwarding utility value according to encounter frequency and the messages are only forwarded to nodes with higher forwarding utility. The algorithm reduces the additional costs arising from the broadcast transmission of messages in epidemic forwarding algorithm. However, PROPHET algorithm only considers the encounter frequency when estimating the forwarding utility, making it difficult to fully describe the impact of the forwarding utility on transmission performance. For these features and shortcomings of PROPHET, a novel hybrid routing algorithm (NHRA) based on the combination of utility and redundancy is proposed. The utility of message forwarding is estimated according to value of the encounter frequency and connectivity duration, and then the maximum number of copies of the message is designated in source node. The forwarding task of the message copies are assigned to the node based on the method of Binary Tree and according to the utility of message forwarding, thus the multi-path parallel transmission can be implemented for message.Then, there is a focus on the reliability of link and a Routing Algorithm Based on Contact-duration Prediction (CPBRA) is proposed. It uses the information of mobile speed, direction and transmission range to predict the contact-duration, and selects the appropriate transmission path according to message's size in order to improve the success rate of communication between nodes, thus reducing the message retransmission, improving the reliability of message transmission and resource utilization. Meanwhile, the algorithm defines the concept of "adjacent nodes". Each node determines whose "adjacent nodes" they are according to the node encounter history information, and diffusion transmission is used to make message be delivered to the "adjacent nodes" of target node quickly, and then the message is only distributed among the "adjacent nodes" of target node to control the network overhead.Finally, our research chooses ONE as simulation platform. This thesis simulates and analyses our proposed routing algorithm by setting different network simulation scenarios to verify the performance. Experimental results show that the proposed routing algorithm NHRA is to some extent suitable for the network scenes with smaller range and relatively stable node-connection. But in large-scale network scenes with node-connection frequently disconnected, CPBRA algorithm shows good transmission performance.