The anomaly detection method based on artificial immune of distributed service

Distributed service is an effective way to solve the massive user services. However, the dynamic combination of services can lead to uncertainty in service, what´ s more, a large number of service´s massive data lead to inefficiency in anomaly detection of service. So it increases the difficulty of the service anomaly detection. This paper inspired by the biological processes of artificial immune recognizing abnormality and propose a method which dynamically detect distributed services abnormal. First of all, we detect abnormal source through numerical differentiation method. Secondly, we draw the ideological of DCA, and through fusion invoking times and average times to calculating danger zone. We achieve the goal of dynamically detecting distributed services abnormal. At last, the experiments verify the feasibility of the method.