Traffic Statistics Determination of Unified Threat Manager to Provide Threat Perception

In the recent time with the increasing rate of data, the security threats on these data have also increased. So, the volume of alerts generated by unified threat manager (UTM) becomes very large. Using conventional methods to analyze a lot of data would drag down the system performance. This paper is a proposal to a system, which will take the UTM logs as the input and analyze these logs using Hadoop Map-Reduce programming mode. Thereby, presenting some threat perceptions by determining the traffic statistics and produce a summarized alarm reports for discovering, predicting and stopping the security threats in the system at a comparatively faster rate. The system would analyze the logs directly at the time of some alerts occurrence and also historical data (logs) from the firewall in order to search for attack signatures (anomaly and patterns) for predicting attacks.