New approach of authentication method based on profiles

The exponential raise in traffic across network has obliged the need to distinguish unauthorized access. The process of identifying an individual usually based on a user name and password, which is the process of giving individuals access to system objects based on their identity. Authentication can involve highly complex and secure methods. This paper proposes an innovative algorithm based on neural networks and fuzzy logic that can be used for easy identification of unauthorized users in a network. The algorithm has been tested and validated and the results have been discussed. The result analysis shows that the algorithm is successful in intrusion detection.