Applied Mechanics and Materials Vols. 263-266

Abstract: Intrusion detection system (IDS) can find the intrusion information before the computer be attacked, and can hold up and response the intrusion in real time. Artificial neural network algorithms play a key role in IDS. The intrusion detection system (ANN) algorithms can analyze the captured data and judge whether the data is intrusion. In this paper we used Back Propagation (BP) network and Radical Basis Function (RBF) network to the IDS. The result of the experiment improve that The RBF neural network is better than BP neural network in the ability of approximation, classification and learning speed. During the procedure there is a large amount of computes. On cloud platform the calculation speed has been greatly increased. So that we can find the invasion more quickly and do the processing works accordingly.

Abstract: Research the problem of intrusion detection in network security. For the defects that existing intrusion detection system can't recognize unknown attacks, to improve the detection efficiency and reduce false alarm rate, this paper basing on V-detector immunity negative intrusion detection algorithm, bring up an adaptive variable radius detector. In the training stage, randomly generate candidate detector with different detection threshold. In the testing stage, according to the size of hole and detection accuracy automatically adjust the radius of the detector. The experiment and the data of KDDCUP1999 show: under the circumstance of the same number of detector, compared with V-detector, this detector has higher coverage, less hole.

Abstract: To identify the malicious nodes timely in wireless sensor networks(WSNs), a cooperation intrusion detection scheme based on weighted k Nearest Neighbour(kNN) is proposed. Given a few types of sensor nodes, the test model extracts the properties of sensor nodes related with the known types of malicious nodes, and establishes sample spaces of all sensor nodes which participate in network activities. According to the known node’s attributes sampled, the unknown type sensor nodes are classified based on weighted kNN. Considering of energy consumption, an intrusion detection system selection algorithm is joined in the sink node. Simulation results show that the scheme has a lower false detection rate and a higher detection rate at the same time, and it can preserve energy of detection nodes compared with an existing intrusion detection scheme.

Abstract: In this paper, we propose a novel key management scheme based on Bezier curves for hierarchical wireless sensor networks (WSNs). The design of our scheme is motivated by the idea that a Bezier curve can be subdivided into arbitrarily continuous pieces of sub Bezier curves. The subdivided sub Bezier curves are easily organized to a hierarchical architecture that is similar to hierarchical WSNs. The subdivided Bezier curves are unique and independent from each other so that it is suitable to assign each node in the WSN with a sub Bezier curve. Since a piece of Bezier curve can be presented by its control points, in the proposed key management scheme, the secret keys for each node are selected from the corresponding Bezier curve’s control points. Comparing with existing key management schemes, the proposed scheme is more suitable for distributing secret keys for hierarchical WSNs and more efficient in terms of computational and storage cost.

Abstract: This paper presents an algorithm by embedding a digital watermark into an original color image. Considering the characteristics of HVS, YUV color space is employed, this algorithm presents a new adaptive digital audio watermarking based on discrete cosine transform. The result of the experiment shows that the proposed algorithm has good robustness against various image processing and attacks, good invisibility and stability.

Abstract: Aiming at the JSTC’s shortcomings, we present an improved JPEG Steganographic method using Syndrome-Trellis Codes with Wetness-Scale in this paper. Firstly, the proposed method excludes the DC coefficients from usable coefficients for information embedding; secondly, when using Syndrome-Trellis Codes to modulate the message bits, it assigns each individual cover element with a special changing cost value provided by our predefined embedding impact model (i.e. wetness-scale), and endeavors to constrain the embedding changes to the low-frequency and middle-frequency DCT coefficients and minimize the embedding impact. The experimental results show that in comparison with JSTC and Jsteg 、F5、MB, the proposed method has stronger ability of anti-steganalysis, and especially when at the embedding rate 0.1, its highest detection accuracy by one of the current best bind steganalysis methods only achieves at 0.5435.

Abstract: Support vector machine (SVM) is suitable for the classification problem which is of small sample, nonlinear, high dimension. SVM in data preprocessing phase, often use genetic algorithm for feature extraction, although it can improve the accuracy of classification. But in feature extraction stage the weak directivity of genetic algorithm impact the time and accuracy of the classification. The ant colony algorithm is used in genetic algorithm selection stage, which is better for the data pretreatment, so as to improve the classification speed and accuracy. The experiment in the KDD99 data set shows that this method is feasible.

Abstract: Because less data can be hidden on the small vector data set , its copyright is very difficult to obtain protection, and the data is vulnerable to attack, so a new double watermarking algorithm was proposed in this paper. Its main features are: 1) it selected feature points from the sequence of Douglas, then embedded watermarking points on both sides of the feature point, and last embedded watermarking points by wavelet transform again; 2) Further more it increased the map graphic deformation control design. The algorithm was applied to experimental data, and the test results showed that the algorithm had good robustness on graphics, rotation, scaling and shifting geometric transformation of points in graph layers. However, due to the smaller vector data set, so the algorithm is robustness on deletion and cropping of points in graph layers.

Abstract: Smartphones providing services like SMS/MMS, emails, online banking, and other applications, are becoming an integral part of our everyday lives. However, the availability of these services and applications provided by smartphones makes them a target for malicious attackers to propagate malware and perform other malicious attacks. To restrain malware propagation, the research advancement of malware containment is summarized in this paper. We provide an overview on malware, which includes the evolution of mobile malware, related concept, infection vectors, and risks. The typical malware containment models are selected to discuss in detail. At last, the current problems and challenges in this field for future work are proposed. This paper indicates that diversity and complexity of mobile malware pose great challenges in modeling malware containment.

Abstract: Under the circumstance of heterogeneous wireless network, the new characteristic of heterogeneous wireless network makes great challenge to the network security. In this paper, we proposed the active defense model based on trust evaluation, introduced sensitive factors and assigned the weight to reflect the trust, through aggregation method, detect the feedback (or recommendation), select the related information, make the model can be dynamic and objective to network status. The simulation result shows that, the proposed model function well, compared with the good and malicious behavior environment.