Khi Lam: Pattern Matching in Network Intrusion Detection Systems

Student's Name: 
Khi Lam
klam@ucsc.edu
Advisor's Name: 
Martine Schlag
Home University: 
University of California, Santa Cruz
AttachmentSize
Image icon Khi.JPG153.39 KB
PDF icon Khi.pdf277.57 KB
Year: 
2005

As the internet population grows and network speeds increase, the internet has became a dangerous place to be without the protection of certain software, such as firewalls and anti-virus. Just by connecting to the internet, your system is vulnerable to attacks from other users taking control over your system or taking private information from your system. As society migrate more services to the internet, such as banking, shopping, and communications, the cost of an intrusion attack on a personal scale can be mere annoyance but on a global scale it can bring down productivity when several servers go down.

We were working to see if there is a more viable way of implementing the software with hardware such as Field Programmable Gate Arrays (FPGA). Software tends to be slow and work linearly but hardware can work at faster speed and are meant for parallel design. But what hardware gains in speed it loses in flexibility and simplicity. We focused only on the most processing intensive part of the software, which is pattern matching.