Dongqi Wang

Research on the Active DDoS Filtering Algorithm Based on IP Flow

Distributed Denial-of-Service flooding attacks against public web servers are increasingly common. It is impossible for the victim servers to work on the individual level of on-going traffic flows. The scheme establishes IP Flow which is used to select proper features for DDoS detection. Five features are analyzed by the experiments. The IP flow statistics is mainly used to allocate the weights for traffic routing by routers. A new algorithm is thus proposed to get efficiently maximum throughput by the traffic filtering, and its feasibility and validity have been verified in real network circumstances. The algorithm shows its advantages that it is with high average detection and with low false alarm and miss alarm. Moreover, it can optimize the network traffic simultaneously with defending against DDoS attack, thus eliminating efficiently the global burst of traffic arising from normal traffic so as to improve greatly the efficiency of servers.

A Modularity Degree Based Heuristic Community Detection Algorithm

A community in a complex network can be seen as a subgroup of nodes that are densely connected. Discovery of community structures is a basic problem of research and can be used in various areas, such as biology, computer science, and sociology. Existing community detection methods usually try to expand or collapse the nodes partitions in order to optimize a given quality function. These optimization function based methods share the same drawback of inefficiency. Here we propose a heuristic algorithm (MDBH algorithm) based on network structure which employs modularity degree as a measure function. Experiments on both synthetic benchmarks and real-world networks show that our algorithm gives competitive accuracy with previous modularity optimization methods, even though it has less computational complexity. Furthermore, due to the use of modularity degree, our algorithm naturally improves the resolution limit in community detection.

AES Key Expansion Algorithm Based on 2D Logistic Mapping

The shortage of original AES key generation and expansion strategy was researched. Based on these researches, the new AES key generation and expansion algorithms are proposed. In these algorithms, two-dimensional Logistic mapping is used to reduce the dependence between sub-keys. Experimental result shows that the security and robustness of the AES sub-keys have been strengthened.

A community finding method for weighted dynamic online social network based on user behavior

Revealing the structural features of social networks is vitally important to both scientific research and practice, and the explosive growth of online social networks in recent years has brought us dramatic advances to understand social structures. Here we proposed a community detection approach based on user interaction behavior in weighted dynamic online social networks. We researched interaction behaviors in online social networks and built a directed and unweighted network model in terms of the Weibo following relationships between social individuals at the very beginning. In order to refine the interaction behavior, level one fuzzy comprehensive evaluation model was employed to describe how closely individuals are connected to each other. According to this intimate degree description, weights are tagged to the prior unweighted model we built. Secondly, a heuristic community detection algorithm for dynamic network was provided based on the improved version of modularity called module density. As for the heuristic rule, we chose greedy strategy and merely fed the algorithms with the changed parts within neighboring time slice. Experimental results show that the proposed algorithm can obtain high accuracy and simultaneously get comparatively lower time complexity than some typical algorithms. More importantly, our algorithm needs no a priori conditions.

A High Capacity Spatial Domain Data Hiding Scheme for Medical Images

This paper presents a spatial steganography scheme with high steganography capacity for medical images. In the proposed scheme, four least significant bits of the cover image are used to hide secret information and a logistic mapping is employed to scramble cover image before embedding operation. In order to achieve minor degradation to cover medical image, the ROI region is excluded manually before embedding operation and an adaptive embedding strategy is employed. Experiment results show that, compared with Fan L’s hiding scheme, our proposed scheme can improve the steganography capacity by 2.25 times, and the maximum peak signal to noise ratio loss and average loss are only 0.78 and 0.45, respectively. In other words, our scheme does not only improve the capacity of the existing method but also can maintain an acceptable quality of the cover image.

A novel trust model for P2P networks

The characteristics of P2P networks which are open, anonymous, and loosely coupled inter-node lead to false resources, malicious evaluation, and syndicates. A related cluster based trust model for P2P networks named RCTrust is presented in this paper. Peers gather in clusters according to their interest similarity and communication history. There are two kinds of trust relationships. In RCTrust, measure of transaction success rate, communications similarity, honesty and timeliness of evaluation are used as parameters to evaluate the credit of peers. Simulative analysis shows that RCTrust model has robustness against malicious attacks and provides a higher rate of successful transaction.

Public Transit Hubs Identification Based on Complex Networks Theory

ABSTRACT Hubs identification is important for the stability and attack tolerance of complex networks. This paper focuses on public transit hubs identification, and it is useful for the optimization, design, and evaluation of public transit systems. Three public transit hubs identification methods are proposed in this paper. The first one is based on comprehensive effects of stations on the distance and the transfer, and the second one is based on preferences of passengers Transfer – Shortest Path, and the third one is based on preferences of passengers Shortest Path – Transfer. Three employed methods are applied in the Shenyang (the capital of Liaoning province of China) bus transit system, and experimental results show that they are available and especially feasible for finding those potential nodes who play key roles in the network but are not commonly regarded as important nodes in practice, and it is beneficial for traffic policy making.

DDoS mitigation in content distribution networks

In this paper, we built up a simulated environment to research the performances of Content Distribution Networks CDN under DDoS attack. In order to research how to enhance CDNs ability of resisting DDoS attacks, a mathematical model of websites deployment was built. Based on the mathematical model, we proposed a website deploying algorithm SD algorithm which can be used by CDN service providers to make website deployment plans. Experimental results showed that deploying websites under the guidance of SD algorithm can significantly reduce the pressure of servers during DDoS attacks.

Centrality-based bipartite local community detection algorithm

As a kind of typical complex network, bipartite network has already received many specialized research. Local community detection is very useful for bipartite network, but it has not yet been systematically researched. To guarantee equivalent partitioning (obtaining coincident community structure starting from arbitrary node) for two category nodes in bipartite network, a centrality-based bipartite local community detection (CBLCD) algorithm is proposed inspired by water surface undulation. The algorithm first gets local centralized subgraphs/subnetworks according to resource allocation index R defined in this paper, and then expands and combines these subnetworks, until the ultimate local communities are detected. The experimental results on some typical network datasets show that the algorithm achieved both good accuracy and stability.

Community detection algorithm based on structural similarity for bipartite networks

Community detection has realistic meaning to the research of network structure. In this paper, we propose a community detection algorithm (BSSCD algorithm) by calculating the similarity of nodes and dividing the nodes with maximum similarity into the same community. In order to accurately calculate the similarity of nodes, we define a novel similarity calculation method which combines Salton index and the improved Logistic function by analyzing the structural characteristics of two types of nodes in bipartite networks and their effects on the density of community. BSSCD algorithm does not require prior knowledge about the number of communities and the result obtained with BSSCD algorithm is very stable. Experiments on real world network datasets show that the similarity calculation method can improve accuracy of similarity calculation and BSSCD algorithm is an efficient method for community detection in bipartite networks.