Research On Spray And Forward Routing Algorithm In Opportunistic Networks

Opportunistic Networks is a wireless self-organizing network, which aims at resolving the data communication in frequent interruption network. In the Opportunistic Networks, it completes the message transmission by exploiting node's mobility, which does not require the source node and destination node within the same communication, the whole process of message transmission is storage, carrying and forward. However in the actual Opportunistic Networks the messages often can not be transferred properly because of the limited node's mobility, limited node storage and network traffic etc. So it will play a great role to improve Opportunistic Networks performance by designing a good efficient routing algorithm in three aspects:storage, carring and forward.This paper firstly gives a detailed overview of Opportunistic Networks, and secondly describes the application scenarios and the research situations of the Opportunistic Networks at home and abroad, finally based on the brief introduction of Opportunistic Networks routing algorithms, we emphatically analyzes the Prophet and the Spray and Focus which are two classical routing algorithms and then propose a new method on our own way. Prophet combines flooding forwarding with forwarding based on Meeting Prediction, and calculates the probability of the nodes according to the frequency of historical encounter between the nodes. The messages are always forwarded to the nodes with high probability during the process of forwarding the messages. The algorithm can reduce the waste of network resources and the channel congestion and other issues caused by flooding algorithm. However, it is irregular during the process of moving the nodes, and the success rate of message transmission is very low. Based on the strategy of controlling copy and mechanisms of the utility value, Spray and Focus greatly improves the transmission performance of the messages, and controlling the number of messages's copy resolves the problem of unlimited proliferation of the messages, and the mechanism of utility value greatly increases the success rate of message transmission, However, the transmission delay of message is increased because it still uses contagious forwarding mechanism to spread the messages'copy.For the inadequacies of the current algorithms, we propose a new routing algorithm MPSFON which integrates four mechanisms:contagious forwarding, controling copy strategy, releasing the information based on the binary tree and calculating the probability of the nodes based on Meeting Prediction. In the process of the messages transmission is divided into two phase:in the spray phase, based on the four mechanisms which MPSFON has proposed the number of messages copies is distributed in the Opportunistic Networks so that messages can achieve multi-path parallel transmission. In the forward phase, calculate the probability of the node to other nodes based on Meeting Prediction, and forward the messages from the node with low probability to the node with high probability according to the probability values until the message encounters the destination node or the message's TTL is expired. The MPSFON proposed aims to reduce network overhead as well as the delay of message transmission and improves the success rate of message delivery.Our experiments are carried out on the platform of the One simulation software. We firstly verify the advantages and disadvantages of MPSFON under three different mobile models by the platform, and then compare MPSFON to the other two classical routing algorithms under the same mobile model. Experimental results show that MPSFON routing algorithm not only has a higher message delivery probability and reduces the number of message forwarded in larger cache space of nodes and nodes dense network.