A Highly Accurate Internal Intrusion Detection and ProtectionSystem Using Time Linked Access Profiles

Abstract

Because most intrusion detection systems and firewalls identify and separate malicious traits that only come from the externalenvironment of the system. It is difficult to differentiate between the actual system users, the internal attackers who access the device.Also, studies claim that these commands can be recognized by analyzing the system calls produced by these commands.This system, therefore, includes an Intrusion Detection and Protection System (IDPS) security system to detect internal attacks on the System Call (SC)using data mining and forensic techniques.However, some attacks have improved their method, further providing security and preventing the user from tracking user access profiles and patterns. In this work,a new technique is proposed to prevents profile attacks by disconnecting access patterns from users for a specified period based onTime Linked Access Profiles (TLAP).To evaluate the performance detection accuracy is calculated and it is improved when compared to the convolution neural network.