Implementation of Multipattern String Matching Accelerated with GPU for Intrusion Detection System

Rangga Nehemia, Charles Lim, Maulahikmah Galinium, Ahmad Rinaldi Widianto

Abstract


Abstract. As Internet-related  security threats continue to increase  in terms of volume and sophistication,  existing  Intrusion Detection System  is also  being challenged to cope with the current Internet  development. Multi Pattern String Matching algorithm accelerated with Graphical  Processing Unit  is  being  utilized  to improve  the packet scanning performance  of the  IDS. This paper implements a Multi Pattern String Matching algorithm, also called Parallel Failureless Aho  Corasick  accelerated with  GPU to  improve the  performance of IDS.  OpenCL library  is  used  to  allow  the  IDS  to  support  various GPU, including  popular GPU such  as NVIDIA and AMD, used  in our research. The  experiment result  shows that the application of Multi  Pattern String Matching using GPU  accelerated platform provides  a speed up,  by up  to 141% in term of throughput compared to the  previous  research.

Full Text: PDF

Refbacks

  • There are currently no refbacks.