Detecting masquerading attack in software and in hardware

Masquerading attack is one of the most dangerous system intrusions. It is very hard to detect this attack, especially in real time for actual network speed rates. In this work the software and hardware approaches for the detection of attacker are analyzed. It is observed that the software solution could not easily reach the necessary rates and the hardware solution is applied in order to achieve the required throughput. Careful analysis of the bioinformatics pattern matching algorithm that is used indicates that exploiting parallelization could improve the performances. Implementing the modified bioinformatics algorithm in the FPGA can increase for several orders the speed in detection process and achieve the necessary results.