Opportunistic Routing Algorithms In Delay Tolerant Networks

DTN(Delay Tolerant Network,DTN)is a new network architecture which can better adapt to interstellar communication,wildlife monitoring,disaster relief and other more challenging environments.Because of the long delay and intermittent connections,the traditional routing algorithms can not be applied directly in the DTN.Therefore,the design of efficient routing algorithms has become one of the hot issues for DTN research community in recent years.This paper briefly introduces the features of DTN,typical applications and current research situation,and reviews the main DTN routing algorithm.It mainly focuses on two types of routing algorithms based on replication and forwarding strategies,respectively.On this basis,to compensate the shortages of SAW,Improved Spray and Waiting routing based on Node Utility(ISAW-NU)is proposed in this paper.The algorithm considers node activity,residual energy and delivery predictability to calculate a node's integrated utility value.In the spray phase,the copies of the message are distributed in proportion to a integrated utility value of node.In the waiting phase,the node that carries a copy of the message no longer delivers message to the destination node by the way of direct transmission,but forwards the message to nodes which have more sufficient energy or a greater integrated utility value during movement.Secondly,to compensate the shortages of Prophet algorithm,Improved Probabilistic Routing based on Multi-information fusion(IPRM)is proposed in this paper.The algorithm calculates the transmission utility using delivery predictability to the destination node,the sociability and the buffer size of node.According to the transmission utility,the node selects a more appropriate relay node to forward the message.In order to further optimize the performance of IPRM,Buffer Management Strategy based on Probability and Diffusion level(BMSPD)is also proposed in this paper.The strategy preferentially forwards messages which have few hop count and high delivery predictability to destination node and preferentially deletes messages which have been forwarded more times,and have small delivery probability and short remaining lifetime.Finally,the performance of the two proposed algorithms and a variety of algorithms in existing papers have been compared by extensive simunaltions based on Opportunistic Network Environment(ONE)simulation platform,and the simulation results verified the effectiveness of the improved algorithm and strategy.