Cyber Times International Journal of Technology & Management

With the coming of Internet age, network security has become a crucial issue for the web applications. There is a large amount of personal, commercial, military, banking, government and other important data on cloud infrastructures worldwide. To protect these data from various attacks, hackers, intrusions, malware, DDoS attacks or disgruntled employees one of the popular components in network security is used, known as Intrusion Detection Systems (IDS). IDS are commonly, software that automates the intrusion detection process and detects possible intrusions. It is one of the significant paths to solve network security difficulties. According to many researches, Artificial Neural Networks can enhance the performance of intrusion detection systems (IDS) when  distinguished with traditional methods. Even so for ANN based IDS, detection precision, especially for user-to-root (U2R), remote-to-local (R2L), Denial of Service (DOS),low-frequent attacks, probing attack and detection stability are still necessary to be increased.

A rule based expert IDS can detect several well-known intrusions with high detection rate, but it is hard to detect new intrusions. Also its signature database needs to be updated frequently and manually. To solve these problems we have proposed FC-ANN based IDS system for distributed cloud environment. Experimental results on the KDD CUP 1999 dataset indicate that our proposed FC-ANN based IDS system is able to detect and prevent the attacks in terms of new attacks (U2R, R2L, Probe and D-DoS) as well as predefined attacks. FC-ANN framework is based on artificial neural networks and fuzzy clustering. Fuzzy clustering with ANN for cloud based IDS provides high stability to detect various types of network intrusions. To detect such signatures or attacks we used the signature Apriori algorithm.