No abstract

Abstract. The objective of this paper is to present an application of learning algorithms to the detection of anomalies in SOA system. As it was not possible to inject errors into the "real" SOA system and to analyze the effect of these errors, a special model of SOA system was designed and implemented. In this system several anomalies were introduced and the effectiveness of algorithms in detecting them were measured. The results of experiments can be used to select efficient algorithm for anomaly detection. Two algorithms: K-means clustering and Kohonen networks were used to detect the unused functionalities and the results of this experiment are discussed. IntroductionWith the growth of computer networking, electronic commerce, and web services, security of networking systems has become very important. Many companies now rely on web services as a major source of revenue. Computer hacking poses significant problems to these companies, as distributed attacks can make their systems or services inoperable for some period of time. As this happens often, an entire area of research, called Intrusion Detection, is devoted to detect these activities. Nowadays many system are based on Service Oriented Architecture (SOA) [1,2] idea. A system based on SOA provides functionalities as a suite of interoperable services that can be used within multiple, separate systems from several business domains. SOA also provides a way for consumers of services, such as web-based applications, to be aware of available SOA-based services. Service-orientation requires loose coupling of services with operating systems, and other technologies that underly applications. SOA separates functionality into distinct units, or services, which developers make accessible over a network in order to allow users to combine and reuse them in the production of applications.The objective of this paper is to present the detection of anomalies in SOA systems by learning algorithms. Related work is presented in section 2 and in section 3, a special model of SOA system which was used in experiments, is presented. In this systems several anomalies were introduced. Four algorithms: Chi-Square statistics, k-means clustering, emerging patterns and Kohonen networks were used to detect anomalies. Detection of anomalies by k-means and Kohonen networks is presented in section 4 and some conclusions are given in section 5.

No abstract