Research On Traffic And Spreading Processes Based On Complex Networks

There are many networks around us which can be seen as complex networks formed by the interaction of individuals with specific functions.Over the past two decades,complex networks have been extensively studied by experts in a variety of fields.In the complex network model of the transportation domain,the problem of packet queuing rules in nodes and the node capacity have not attracted the attention of scholars.The queuing rules of packets can not only follow the first-in-first-out rule,but also follow the last-in-first-out rule and random-in-random-out rule.In addition,in the past research,the node capacity is usually infinite by default,however,the capacity of most networks in our reality is limited.Therefore,based on the condition of limited network capacity,for a network that adopts last-in-first-out and last-in-first-out rules,what characteristics will the traffic congestion phenomena show when a network model and routing strategy consistent with the first-in-first-out rule is adopted? In a network with different node capacities,what is the difference in the state of epidemic spreading under different routing strategies? Based on the related knowledges of complex network,the above questions are deeply studied through theoretical analysis,and the following main tasks are carried out:(1)The complex network model and the epidemic spreading model based on it have been studied in this paper.The concept,the history and the topology of complex networks are explained in detail,and the topology of four types of typical static networks and one type of dynamic network have been studied.Three basic epidemic spreading models based on complex networks have been discussed in detail,and the epidemic thresholds of three typical epidemic spreading models in scale-free networks have been studied.The usability of the model have been proposed.(2)The influence of queuing rules on traffic congestion in dynamic networks has been studied.A routing strategy based on the physical distance between two nodes and node load states has been proposed.In a single-layer dynamic network,packet queuing rules of first-in-first-out,last-in-first-out and last-in-first-out are adopted respectively.Through simulation experiments,the effects of packet generation rate,tunable parameter of routing strategy and node capacity on network transmission state have been analyzed,and these influences have been compared.Moreover,a phenomenon similar to Braess’ s paradox under these three queuing rules has been observed.(3)The influence of routing strategy on traffic congestion and epidemic spreading in BA scale-free network has been studied.The packet storage capacity of nodes related to the degree of nodes has been proposed.The shortest route strategy and effective routing strategy as well as the limited node sending capacity are respectively adopted on the BA network.Through the simulation experiment,the influence of the probability of package generation has been analyzed.The initial infection source ratio based on two routing path strategies on the network transmission capacity and packet storage rate has been analyzed.It is found that the traffic congestion of central nodes is helpful to suppress the spreading behavior.Moreover,we found that the number of initial infectious sources does not affect the epidemic threshold.