Publications

In this paper, we abstract a kind of stochastic processes from evolving processes of growing networks, this process is called growing network Markov chains. Thus the existence and the formulas of degree distribution are transformed to the corresponding problems of growing network Markov chains. First we investigate the growing network Markov chains, and obtain the condition in which the steady degree distribution exists and get its exact formulas. Then we apply it to various growing networks. With this method, we get a rigorous, exact and unified solution of the steady degree distribution for growing networks.

In this paper, we investigate random walks in a family of small-world trees having an exponential degree distribution. First, we address a trapping problem, that is, a particular case of random walks with an immobile trap located at the initial node. We obtain the exact mean trapping time defined as the average of first-passage time (FPT) from all nodes to the trap, which scales linearly with the network order $N$ in large networks. Then, we determine analytically the mean sending time, which is the mean of the FPTs from the initial node to all other nodes, and show that it grows with $N$ in the order of $N \ln N$. After that, we compute the precise global mean first-passage time among all pairs of nodes and find that it also varies in the order of $N \ln N$ in the large limit of $N$. After obtaining the relevant quantities, we compare them with each other and related our results to the efficiency for information transmission by regarding the walker as an information messenger. Finally, we compare our results with those previously reported for other trees with different structural properties (e.g., degree distribution), such as the standard fractal trees and the scale-free small-world trees, and show that the shortest path between a pair of nodes in a tree is responsible for the scaling of FPT between the two nodes.

No abstract is provided for this article.

The current Internet is dramatically suffering the Distributed Denial of Service (DDoS) attacks, in which the perpetrator maliciously makes network resource unavailable to its intended users by temporarily or indefinitely disrupting the services of a host connected to the Internet. In this paper, we investigate an Internet transmission control protocol/active queue management (TCP/AQM) router subject to DDoS attacks. We utilize the time delay control theory to analyze the dynamics of the congestion control windows, and the queues at the router. We derive some explicit conditions under which the TCP/AQM system under DDoS attacks is asymptotically stable. We discuss the convergence of the queue lengths in the router. Our results suggest that, if the network parameters in the TCP window updating, and control parameters in the AQM algorithm satisfy certain conditions, the TCP/AQM system is stable, and its queue lengths can converge to any given target. This result is important, and promising in terms of applications in that, when the DDoS attacked traffic is differentiated from the legitimate traffic, one is able to choke the DDoS attacks by limiting their rates, and then to improve the bandwidth usage of the normal flows. We illustrate the theoretical results using the network simulation platform <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$ns2$</tex-math></inline-formula> , and demonstrate that the controlled network can achieve good performance, enhancing the Internet robustness, and performance against DDoS attacks.

Most existing flocking algorithms assume one single virtual leader and rely on information about both relative positions and relative velocities among neighboring agents. In this paper, we investigate the problem of controlling a flock of mobile autonomous agents to follow multiple virtual leaders using only position information in the sense that agents with the same virtual leader asymptotically attain the same velocity and track the corresponding virtual leader based only on position measurements. We propose a flocking algorithm, under which every agent asymptotically attains its desired velocity, collision between agents can be avoided, and the final tight formation minimizes all agents' global potentials. We show one simulation to verify and illustrate the theoretical results.

Measuring the influence of scientists and their activities on science and society is important and indeed essential for many studies. Despite the substantial efforts devoted to exploring the influence's measures and patterns of an individual scientific enterprise, it remains unclear how to quantify the mutual impact of multiple scientific activities. This work quantifies the relationship between the scientists' interactive activities and their influences with different patterns in the AMiner dataset. Specifically, inflation treatment and field normalization are introduced to process the big data of paper citations as the scientist's influence, and then the evolution of the influence is investigated for scientific activities in the citation and cooperation patterns through the Hawkes process. The results show that elite scientists have higher individual and interaction influences than ordinary scientists in all patterns found in the study, with permutation tests verifying the significance of the new findings. Moreover, the study compares the patterns found in two largest disciplines, i.e., <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">STEM</i> and <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Humanities</i> , revealing the higher value of individual influence in <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">STEM</i> than in <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Humanities</i> . Furthermore, it is found that the opposite trend of <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">STEM</i> and <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Humanities</i> in the cooperation pattern suggests different cooperation habits of scientists in different disciplines. Overall, this investigation provides a feasible approach to addressing the scientific influence issue and deepening the quantitative understanding of the mutual influence of multiple scientific activities in science and society.

Network robustness is critical for various societal and industrial networks again malicious attacks. In particular, connectivity robustness and controllability robustness reflect how well a networked system can maintain its connectedness and controllability against destructive attacks, which can be quantified by a sequence of values that record the remaining connectivity and controllability of the network after a sequence of node- or edge-removal attacks. Traditionally, robustness is determined by attack simulations, which are computationally very time-consuming or even practically infeasible. In this paper, an improved method for network robustness prediction is developed based on learning feature representation using convolutional neural network (LFR-CNN). In this scheme, higher-dimensional network data are compressed to lower-dimensional representations, and then passed to a CNN to perform robustness prediction. Extensive experimental studies on both synthetic and real-world networks, both directed and undirected, demonstrate that 1) the proposed LFR-CNN performs better than other two state-of-the-art prediction methods, with significantly lower prediction errors; 2) LFR-CNN is insensitive to the variation of the network size, which significantly extends its applicability; 3) although LFR-CNN needs more time to perform feature learning, it can achieve accurate prediction faster than attack simulations; 4) LFR-CNN not only can accurately predict network robustness, but also provides a good indicator for connectivity robustness, better than the classical spectral measures.

The football robotics competition has become one of the typical systems used in multi-robot research to study robot control, artificial intelligence and task allocation. The design of a football robot is a more integrated research module including robot design, visual information processing, and robot motion control. In this paper, a monocular global camera is used on the vision acquisition module to capture information within the football field in real time. Edge detection and feature point extraction are performed on the images after color recognition processing to obtain the position coordinates of the key elements of the football field, the football and the robot, thus significantly reducing the amount of information required for image processing. In addition to this, this paper builds Petri network models for different role robots to achieve different motion state control and analyses the temporal error of the robot's motion. The image processing approach and Petri net modelling used in this paper provide research directions for future indoor multi-robot localization approaches and task allocation.