Traffic dynamics based on a traffic awareness routing strategy on scale-free networks

By incorporating local traffic information into the shortest path routing strategy, we numerically investigate the effectiveness of the traffic awareness routing strategy for scale-free networks with different clustering. In order to characterize the efficiency of the packet-delivery process, we introduce an order parameter and an average transmission time that allow us to measure the network capacity by the critical value of phase transition from free flow to congestion. Compared with the shortest path routing protocol, the network capacity is greatly enhanced by the traffic awareness routing strategy. We also find that there exists an optimum value for the tunable parameter in the congestion awareness strategy. Moreover, simulation results show that the more clustered the network, the less efficient the packet-delivery process.     

Decorrelation of networked communication flow via load-dependent routing weights

Clever assignments of link weights are able to change communication routes in such a way that loads are distributed almost evenly across a network. This is achieved by weight assignments based on the link load. As demonstrated for scale-free as well as synthetic Internet networks, they decorrelate the loads of the nodes and links from the network structure and increase the transport capacity of the network. For various Internet scans the increase of transport capacity amounts to a factor of around five when compared to shortest-path routing.     

Information traffic in scale-free networks with fluctuations in packet generation rate

We study the information traffic in scale-free networks where the information generation rate varies with time as a periodic function. We observe that when the fluctuation in packet generation rate increases, the average transit time increases and network performance degrades. In order to improve the transportation efficiency in this situation, we propose a new routing method called mixed routing. It operates in two modes: (1) when the packet generation rate is small, the shortest paths are used to deliver the packets to the destination; (2) when the packet generation rate is large, the traffic loads in central nodes are redistributed to other non-central nodes, using the so-called efficient routing method. We find that the time shifting between the two modes is very critical for the routing performance. Consequently, we provide an efficient method to determine the critical times to shift the routing modes for achieving good network performance.     

Detrended Fluctuation Analysis of Traffic Data

Different routing strategies may result in different behaviour of traffic on internet. We analyse the correlation of traffic data for three typical routing strategies by the detrended fluctuation analysis （DFA） and lind that the aregree of correlation of the data can be divided into three regions, i.e. weak, medium, and strong correlation. The DFA scalings are constants in both the regions of weak and strong correlations but monotonically increase in the region of medium correlation. We suggest that it is better to consider the traffic on complex network as three phases, i.e. the free, buffer, and congestion phase, than just as two phases believed before, i.e. the free and congestion phase.     

Macro-Behaviour of Agents＇ Opinion under Influence of an External Field

We propose a model to study the evolution of opinion under the influence of all external field on small world network. The maero-behaviour of agents＇ opinion and the relative change rate as time elapses are studied. The external field is found to play an important role in making the opinion s（t） balance or increase, and without the influence of the external field, the relative change rate γ（t） shows nonlinear increasing behaviour as time runs. What is more, this nonlinear increasing behaviour is independent of the initial condition, the strength of the external field and the time that we cancel the external field. The results may reflect some phenomena in our society, such as the function of the macro-control in China or the mass media in our society.     

Is the topology of the Internet network really fit to sustain its function?

The Internet is one of the most interesting realizations of a “complex” network. As a non-supervised growing object, it allows the study of the selective pressure which drives the network to assume its current structure. The DIMES and the ROUTEVIEWS projects are ongoing projects aimed at evaluating the topological structure of the Internet (at the Autonomous System or AS grain-level) on the basis of different types of measurements. The topological analysis of the networks produced by the two projects has allowed us to infer a growth mechanism which has been used to build up synthetic networks with similar properties. These networks have been used as test-beds for the implementation of a model of traffic dynamics, with the aim of assessing the ability of the Internet’s topology to support the basic actions for data traffic handling. Results have been compared with those obtained by using a random network of similar size. The effects of some structural perturbations (arcs and nodes’ removal, traffic localization) have been also evaluated in terms of the induced variations of the network’s efficiency. The resulting scenario is consistent with the hypothesis that the structure of the Internet is only partially fit to host communication processes and that the intelligence of the TCP/IP protocol is partly needed to overcome some “structural” deficiencies.     

Geographical Effects on Complex Networks

We investigate how the geographical structure of a complex network affects its network topology, synchronization and the average spatial length of edges. The geographical structure means that the connecting probability of two nodes is related to the spatial distance of the two nodes. Our simulation results show that the geographical structure changes the network topology. The synchronization tendency is enhanced and the average spatial length of edges is enlarged when the node can randomly connect to the further one. Analytic results support our understanding of the phenomena.     

Cross-impact balances: Applying pair interaction systems and multi-value Kauffman nets to multidisciplinary systems analysis

Cross-impact analysis is the name for a familiar method for multidisciplinary systems analysis in social sciences and management sciences, especially in technology foresight, technology assessment and scenario planning. A recently proposed form of cross-impact analysis, CIB, may be of interest for physicists, sociophysicists and complex network researchers because the CIB concept reveals considerable relations to some concepts of these research fields.This article describes the basics of CIB analysis framework, its applications in the social sciences, and its relations to the equilibrium points of pair interaction systems, random graphs, and generalized Kauffman nets.Therefore CIB can be seen as a merger of concepts originating in utterly different scientific fields. This may prove to be fruitful for both sides: For sociophysicists as an example of the application of complex network concepts in the social sciences and for cross-impact practitioners as a source of theoretical insights in the background of their tool.     

Detect overlapping and hierarchical community structure in networks

Clustering and community structure is crucial for many network systems and the related dynamic processes. It has been shown that communities are usually overlapping and hierarchical. However, previous methods investigate these two properties of community structure separately. This paper proposes an algorithm (EAGLE) to detect both the overlapping and hierarchical properties of complex community structure together. This algorithm deals with the set of maximal cliques and adopts an agglomerative framework. The quality function of modularity is extended to evaluate the goodness of a cover. The examples of application to real world networks give excellent results.     

Generalized Projective Synchronization between Two Complex Networks with Time-Varying Coupling Delay

Generalized projective synchronization （GPS） between two complex networks with time-varying coupling delay is investigated. Based on the Lyapunov stability theory, a nonlinear controller and adaptive updated laws are designed. Feasibility of the proposed scheme is proven in theory. Moreover, two numerical examples are presented, using the energy resource system and Lü＇s system [Physica A 382 （2007） 672] as the nodes of the networks. GPS between two energy resource complex networks with time-varying coupling delay is achieved. This study can widen the application range of the generalized synchronization methods and will be instructive for the demand supply of energy resource in some regions of China.     

A Non-Stationary Poisson Model for the Scaling of Urban Traffic Fluctuations

We investigate the traffic flow volume data on the time dependent activity of Beijing＇s urban road network. The couplings between the average flux and the fluctuations on individual links are shown to follow certain scaling laws and yield a wide variety of scaling exponents between 1/2 and 1. To quantitatively explain this interesting phenomenon, a non-stationary Poisson arriving model is proposed. The scaling property is interpreted as the result of the time- variation of the arriving rate of flux over the network, which nicely explicates the effect of aggregation windows, and provides a concise model for the dependence of scaling exponent on the external/internal force ratio.     

A unified model for Sierpinski networks with scale-free scaling and small-world effect

In this paper, we propose an evolving Sierpinski gasket, based on which we establish a model of evolutionary Sierpinski networks (ESNs) that unifies deterministic Sierpinski network [Z.Z. Zhang, S.G. Zhou, T. Zou, L.C. Chen, J.H. Guan, Eur. Phys. J. B 60 (2007) 259] and random Sierpinski network [Z.Z. Zhang, S.G. Zhou, Z. Su, T. Zou, J.H. Guan, Eur. Phys. J. B 65 (2008) 141] to the same framework. We suggest an iterative algorithm generating the ESNs. On the basis of the algorithm, some relevant properties of presented networks are calculated or predicted analytically. Analytical solution shows that the networks under consideration follow a power-law degree distribution, with the distribution exponent continuously tuned in a wide range. The obtained accurate expression of clustering coefficient, together with the prediction of average path length reveals that the ESNs possess small-world effect. All our theoretical results are successfully contrasted by numerical simulations. Moreover, the evolutionary prisoner’s dilemma game is also studied on some limitations of the ESNs, i.e., deterministic Sierpinski network and random Sierpinski network.     

Local Events and Dynamics on Weighted Complex Networks

We examine the weighted networks grown and evolved by local events, such as the addition of new vertices and links and we show that depending on frequency of the events, a generalized power-law distribution of strength can emerge. Continuum theory is used to predict the scaling function as well as the exponents, which is in good agreement with the numerical simulation results. Depending on event frequency, power-law distributions of degree and weight can also be expected. Probability saturation phenomena for small strength and degree in many real world networks can be reproduced. Particularly, the non-trivial clustering coefficient, assortativity coefficient and degree-strength correlation in our model are all consistent with empirical evidences.     

Scale-free networks by super-linear preferential attachment rule

A network growth model with geographic limitation of accessible information about the status of existing nodes is investigated. In this model, the probability Π(k) of an existing node of degree k is found to be super-linear with Π(k)∼k^α and α>1 when there are links from new nodes. The numerical results show that the constructed networks have typical power-law degree distributions P(k)∼k^−γ and the exponent γ depends on the constraint level. An analysis of local structural features shows the robust emergence of scale-free network structure in spite of the super-linear preferential attachment rule. This local structural feature is directly associated with the geographical connection constraints which are widely observed in many real networks.     

Diffusion-Based Recommendation in Collaborative Tagging Systems

Recently, collaborative tagging systems have attracted more and more attention and have been wlaely appnea in web systems. Tags provide highly abstracted information about personal preferences and item content, and therefore have the potential to help in improving better personalized recommendations, We propose a diffusion- based recommendation algorithm considering the personal vocabulary and evaluate it in a real-world dataset： Del.icio.us. Experimental results demonstrate that the usage of tag information can significantly improve the accuracy of personalized recommendations.     

Dynamics of traffic networks: From microscopic and macroscopic perspectives

Considering the microscopic characteristics (vehicle speed, road length etc.) of links and macroscopic behaviors of traffic systems, we derive the critical flow generation rate in scale-free networks. And the dynamics of traffic congestion is studied numerically in this paper. It is shown that the queue length increases with microscopic characteristics of links. Additionally, the critical flow generation rate decreases with increase of the network size N, maximum speed v_max and parameter τ. The significance of this finding is that, in order to improve the traffic environment, both the local information for the single link and behaviors of the whole network must be analyzed simultaneously in a traffic system design.     

A class of improved algorithms for detecting communities in complex networks

Detecting communities in complex networks is of considerable importance for understanding both the structure and function of the networks. Here, we propose a class of improved algorithms for community detection, by combining the betweenness algorithm of Girvan and Newman with the edge weight defined by the edge-clustering coefficient. The improved algorithms are tested on some artificial and real-world networks, and the results show that they can detect communities of networks more effectively in both unweighted and weighted cases. In addition, the technique for improving the betweenness algorithm in this paper, thanks to its compatibility, can directly be applied to various detection algorithms.     

Traffic dynamics in scale-free networks with limited packet-delivering capacity

We propose a limited packet-delivering capacity model for traffic dynamics in scale-free networks. In this model, the total node’s packet-delivering capacity is fixed, and the allocation of packet-delivering capacity on node i is proportional to , where k_i is the degree of node i and ? is a adjustable parameter. We have applied this model on the shortest path routing strategy as well as the local routing strategy, and found that there exists an optimal value of parameter ? leading to the maximal network capacity under both routing strategies. We provide some explanations for the emergence of optimal ?.