| Delay-Tolerant Network (DTN) is a new kind of wireless network concept, whichguarantees the communications requirements of a set of networks characterized withhigh link delays, weak signal strength, large bidirectional data-rate asymmetries, andintermittent connectivity, such as deep space network, Battlefield communication net-work etc. Due to the differences between the DTN’s characteristics, protocol system andtraditional network’s, the main work of the DTN research is to design the protocols thatcan tolerant delay and disruption.Firstly, a routing algorithm based on gradient and fuzzy neural network decisionfor DTN is proposed in this paper. It improves the vector for network describing byadopting the node’s own information and the link state information between nodes,which enables a comprehensive description of the network. The algorithm obtains ac-curate measurement for routing decisions by combining the dynamic average of the lim-ited historical information with accurate prediction and maintaining the components ofthe vector adaptively. In addition, fuzzy radical basis function neural network is em-ployed for routing decision, which leads to makes the routing decision-making processto be intelligent. The packet forwarding efficiency is improved by guiding the forwarddirection along the gradient of multi-hop packet transmission probability of success.Through simulative analysis, the algorithm proposed is found to achieve better per-formance than Epidemic and Context-Aware Routing algorithms under the same net-work conditions. Secondly a file delivery protocol based on multi-file united delivery isproposed too. It unites the processes of each files’ delivery to let files utilize the frontfiles’ RTD (Round Trip Delay). This protocol promotes the link efficiency and reducestransmission time at limited cost when many files need to be delivered. Through theo-retical and simulation analysis, this protocol gets performance boost relative to CFDP. |
PDF Full Text Request