Hua Yang’s research while affiliated with Sichuan University and other places

The goal of this research is to detect anomaly of the data from a network by an artificial immune system. A new immune-based approach with feedback is proposed to adapt to the complex dynamic network environment by using the receptor density algorithm. The results of the algorithm demonstrate that the new approach lowers the false positive and thus enhances the efficiency and reliability for anomaly detection without an increase in complexity.

In the area of computer security, Intrusion Detection (ID) is a mechanism that attempts to discover abnormal access to computers by analyzing various interactions. There is a lot of literature about ID, but this study only surveys the approaches based on Artificial Immune System (AIS). The use of AIS in ID is an appealing concept in current techniques. This paper summarizes AIS based ID methods from a new view point; moreover, a framework is proposed for the design of AIS based ID Systems (IDSs). This framework is analyzed and discussed based on three core aspects: antibody/antigen encoding, generation algorithm, and evolution mode. Then we collate the commonly used algorithms, their implementation characteristics, and the development of IDSs into this framework. Finally, some of the future challenges in this area are also highlighted.

The prediction of an image quality metric (IQM) should be consistent with subjective human evaluation. Since the human visual system (HVS) is critical to visual perception, modeling of the HVS has been regarded as the most suitable way to achieve perceptual quality predictions. In fact, sparse coding, which is equivalent to independent component analysis (ICA) can provide a very good description of the receptive fields of simple cells in the primary visual cortex which is the most important part of the HVS. Inspired by this fact, a quality metric called sparse feature fidelity (SFF) is proposed for full-reference image quality assessment (IQA). It is based on the fact that images are transformed into sparse representations in the primary visual cortex. The proposed method is based on sparse features that are acquired by a feature detector which is trained on samples of natural images by an ICA algorithm. Moreover, two strategies are designed to simulate the properties of the visual perception: visual attention and visual threshold. The computation of SFF has two stages: training and fidelity computation, moreover, the fidelity computation consists of two components: feature similarity and luminance correlation. The component of feature similarity measures the structure differences between two images, while the luminance correlation evaluates brightness distortions. SFF also reflects the chromatic properties of the HVS, and it is very effective for color IQA. Experimental results on five image databases show that SFF has a better performance in matching subjective ratings when compared with the leading IQMs.

In this paper, a method for accelerating registration of partially overlapping images is presented. Image registration is a fundamental problem in image processing and computer vision. Numerous studies have been done on this subject and can satisfy most of engineering application requirements. However, the registration processing is very time-consuming when the images are very large. So, time-saving methods are desirable. The proposed method, which is based on a feature-based method, concerns the computational cost of registration of partially overlapping images. The feature-based method generally contains four steps--features detecting, features matching, estimation of the homography and image warping and blending. The new method includes two stages. In the first stage, overlapping areas are rapidly estimated from the low-resolution correspondences of candidate images by a SIFT-based method. In the second stage, a homography between the image pairs is calculated by the same SIFT-based method, which involves only the overlapping areas between a target image and a reference image. Then the target image is warped and blended with the reference image. The experimental results demonstrated that our proposed method can reduce the computational cost to 10%similar to 30% of that of the SIFT-based homographic estimation (SHE) methods with little compromise in accuracy. (C) 2011 Published by Elsevier Ltd. Selection and/or peer-review under responsibility of [CEIS 2011]

Cloud computing technology not only provides us powerful computing, on-demand service, rapid elasticity, but also possible great destruction by the internet criminal accordingly. This prompts us to consider the cloud-based worm propagation problem. We set up the analytical model through a highly abstract network environment and achieve the overall characteristics of the worm research purposes. In this paper, we firstly analyze the factors affect worm scan and propagation in the cloud and put forward a novel cloud-based worm model: the MapReduce Divide-and-Conquer model (MRDC). Secondly, we analyze the architecture and performance of MRDC worm contrast to Code Red, the hit-list worm, and flash worm, etc. the simulation shows that the MRDC worm significantly improves the worm propagation. And finally, we discuss some threat trends of cloud-based worm propagation and some possible solutions. Our simulation shows that the MRDC worm propagate much faster than the other worms that the MRDC worm can scan and infect entire IPv4 space in no more than 10 minutes, the perfect MRDC worm can infect 360,000 vulnerable machines in no more than 1 second and all vulnerable machines in the entire IPv4 space in no more than 10 seconds.