Opportunistic Network Forwarding Technique Based On Node Cosine Similarity Of Nodal Positions

With the popularity of mobile devices,a large number of handheld mobile devices as nodes can form an opportunistic network.The traditional communication is to complete the communication through connecting between the end-to-end,and the links between the nodes are not stable,and the data cannot be transmitted normally by relying on the traditional communication.This problem can be solved by the opportunistic forwarding mode of data,and mobile devices carried by people can interact with each other to realize data transmission.Therefore,research on data forwarding technology is very necessary,and the activity characteristics of nodes can be predicted through the information exchanged between nodes,so as to improve the transmission success rate and reduce the forwarding cost.In practical applications,since the node mobility is typically complex,and the location of the destination node is constantly changing,it is difficult to obtain the real-time locations of nodes.If the data dissemination is decided by the distance between nodes,and then the distribution of nodes holding copies may become extremely uneven.Under such case,how to improve the delivery ratio with a minor dissemination cost and a shorter dissemination delay become the vital issue.In our work,some historical information(node encounters,length of meeting period,and stability of node relationship)is used to calculate the node's forwarding utility value,and the node with the largest utility value is selected as the first forwarding node.Then,the forwarding utility and the cosine similarity method are combined to select other forwarding nodes.In this thesis,an Opportunistic Network forwarding algorithm based on Node Cosine Similarity(ONNCS)is proposed.ONNCS tries to make the data holders be evenly distributed,so that the data packets can be forwarded to the destination node as soon as possible.The simulation experiment results show that ONNCS can achieve a high delivery ratio and a low energy consumption.Moreover,an Android prototype system where point-to-point communication is implemented based on the Wi-Fi P2 P is designed and realized.It integrates some functions such as nodes searching,connection establishing and data packet exchanges with surrounding nodes.Some test results regarding the system performance are collected,performing real-life experiments on related algorithms and the prototype can provide some supports for the future research in the field of opportunistic networks.