Research On Malicious Code Detection Technology Based On Cloud Computing

The network era has not only brought a variety of convenience of life,but also brought the outbreak of malicious code growth. The report notes that the number and types of malicious code are growing rapidly, Anti-killing methods of malicious code are also progressing, encrypting the shell has become one of the necessary technologies of malicious code, anti-virus software often needs to get the dynamic behavior of malicious code through virtual machine. The increase in the difficulty of malicious code detection makes the user’s personal computer more difficult to cope with its attack, cloud detection becomes the current hot detection method, the development of cloud computing technology makes the development of cloud-based malicious code detection technology possible.At present, the domestic and foreign research focused on serial technology, effective use of existing cloud computing technology is not enough, detection efficiency is not high, a lot of high accuracy but the arithmetic did not get enough development. Based on this, this paper studies the existing malicious code detection technology, and puts forward the malicious code detection technology based on cloud computing. The contributions of this paper are as follows:(1) Details of the development of the Internet and the status quo of security, as well as domestic and international research status of malicious code detection technology.(2) Through the study of the existing malicious code detection technology and cloud computing technology, this paper proposes a malicious code detection technology based on cloud computing. The extraction method of malicious code runtime system calls the API sequence as a feature vector trained with DBScan algorithm to cluster the training samples, and then use the SVM algorithm to train various types of classification model, detection will be detected when the training data is divided into clustering each class, and then SVM algorithm to detect. In this paper, the detection algorithm is transplanted to the cloud computing platform Hadoop, and the serial algorithm is optimized to the parallel algorithm.(3) The design and implementation of the overall module of the malicious code detection technology based on cloud computing, and in-depth study and implementation of the key technologies.(4) Select the training samples and test samples of malicious code detection scheme based on cloud computing experiments, the experimental results show that the proposed method can effectively detect the malicious code and greatly improve the detection efficiency.