Combination of RBF Algorithm and Filters Applied to Prevent Attacks in a Web Server

Intrusion detection is a critical process in network security and of course intrusion prevention is needed to secure the system. The proposed work of adaptive smart web server monitors the intrusions or vulnerabilities of the computer network and eliminates it. The proposed scheme provides data authentication, confidentiality in computer networks. The various study and development of intrusion detection had been done using neural algorithms like Radial Basis Function (RBF) with its trained sets of rules. We have developed a smart web server model with finger print authentication and filtering techniques like applet filter, virus filter, malware pop up filter, onload/on unload filter, attack filter. Confidentiality is very important for the information related to militaries, banking and medical databases etc. In this proposed paper, bio-metric fingerprint authentication is given to smart web servers FTP access to prevent the attackers from the malicious activities. Few attacks like IP sequence prediction attacks, IP fragmentation attacks and Back End Server attacks have been detected in the log file entries. This idea of filtering and parsing the web content before it is published in the WWW (world wide web) will be the best method for the data authentication, user authentication and content authentication. By implementing the same standards of authentication to the entire web servers, we can prevent the hacking possibilities.