Research On Opportunistic Network Routing Algorithm Based On User Interest And Trust Relationship

In recent years,with the rapid development of wireless communication technology and the universal popularization of mobile communication devices(smartphones,PDAs,etc.),people have used short-distance wireless communication interfaces such as Bluetooth or WiFi to realize data transmission and information sharing through self-organization method.That makes it possible to communicate in a network environment that does not have basic communication facilities,which has led to a new type of mobile self-organizing network-Opportunistic Networks.This new type of network has huge application prospects in many areas such as vehicular network,remote area network communications,post-disaster emergency rescue and media information sharing.The opportunistic network is a kind of wireless self-organizing network that does not require a complete link between the source node and the destination node,and it uses the encountering opportunities created by the moving nodes to realize communication with the "store-carry-forward" mode.The network communication employs "storage-carry-forward" routing strategy.Due to the dynamic topology changes of the opportunistic network,the traditional network protocols are no longer applicable.Therefore,how to make better use of the opportunistic connections between users and conduct the efficient and reliable transmission and reliable is a major issue for research workers.The fundamental solution to this problem lies in the ability to design a reasonable node mobility model and an efficient routing algorithm.Firstly,this paper introduces the background significance of research opportunistic network,the current research status at home and abroad as well as the main application fields of the opportunistic network.Secondly,we introduces the relevant theoretical knowledge of the opportunistic network,and summarizes and analyze the existing node mobility model and routing algorithm.Although existing routing algorithms can achieve better performance under certain network conditions,they do not make better use of interest features and mobility characteristics of nodes,and they do not solve the problems of selfish behavior due to limited resources.Therefore,this paper carries out the following research work based on these two aspects.(1)User interest based routing algorithm.Firstly,we analyze the characteristics of the user's mobile behavior through social network theory,and we find that people's social relations have a certain dependen.ce and their movements show a small-world phenomenon,they often connects together due to some common interests and social connections,forming a moving group,which also known as the community.Based on this,this study periodically and dynamically divides the network into multiple interest communities through the connection strength and interest similarity of nodes.The message transmission process is divided into the process within interest community and the process between interest communities.Then,we introduce the corresponding transmission strategy in detail.Current nodes decides whether to forward the message according to the utility value of the meeting node and the current message within interest community,makes the message moves in the direction of more efficient value with the destination node,and it select the next node to send the message according to the probability of visiting the community where the destination node is located between interest communities.Finally,the simulation results show that the proposed algorithm has a significant improvement in network performance,a better message delivery rate,a lower average message transmission delay,and a higher usage efficiency of network resources.(2)Trust relationship based routing algorithm.Firstly,we analyzes the causes of the selfish behavior between nodes in opportunistic network,and points out the drawbacks in related works.After considering the direct and indirect trust relationships between nodes,this algorithm establishes the trust model according to the interaction information and the recommendation information of the trusted neighbor nodes,and a trust relationship matrix is established according to the overall trust degree.Then,we use the naive Bayes method to classify the friend relationships according to historical interaction information between nodes.During the process of message transmission,on the premise of ensuring that the encountering node is in a friend node with the destination node,a node with a higher trust degree of the destination node is selected as a relay node for forwarding the message through the established trust relationship matrix and distribute the message copies dynamically according to the trust relationship which makes the message moves in the direction of increasing degree of trust.Finally,the simulation results show that the proposed algorithm has a better message delivery rate,a lower message transmission delay,a smaller system overhead rate and packet loss number in the presence of self-private nodes in the network.