Due to the enormous demand for networking services, the performance and security of information has to be improved. To provide information security, numerous cryptographic algorithms were proposed by various researchers, out of which RSA algorithm is one the most popular algorithm. RSA algorithm uses linear congruence method which restricted the operation to specific class of values. RSA algorithm needs exponential time for decryption of message. By extending the RSA algorithm using congruence class and selecting the key in random, the security of algorithm can be increased. Higher the congruence class index, higher will be its level of security. For each of the congruence class element, complexity of algorithm will be same but there will be increase in the level of security. The basic idea behind this implementation is that by converting the given linear congruence into congruence class and solving them algebraically, actual information can be produced. This paper contains the comparison between linear congruence and congruence class using RSA algorithm. Finally we contend that, due to congruence class implementation, the complexity of algorithm will remain same as that of regular RSA algorithm with enhancement in information security by random congruence class key selection.Copy Right, IJAR, 2016,. All rights reserved.

Due to the growth in prominence of Web, there is a need for proficient system administration. Network visibility becomes very crucial for traffic engineering and network management. A large number of users demands varied information at a given time. By identifying the users that demand same type of information and clustering them into different groups, the Internet accessibility and resource utilization can be improved. The most popular solutions for network management are Deep Packet Inspection algorithm, In-Depth Packet Inspection algorithm and some related statistical classification technologies. All these solutions depend on the availability of a training set. Supervised (classification) and unsupervised (clustering) algorithms are used for identification of the network traffic. Network traffic analysis always depends on various parameters such as the data to be searched, the time of searching, available bandwidth, number of accessing users, architecture of the network system, etc. For simplicity, the type of data and the data rate was considered for this implementation. Due to clustering, automatic identification of the classes of traffic was achieved. Since clustering technique is used for group processing of information, group signature techniques is being applied here for secured data processing.