The world has seen rapid advances in science and technology in the last two decades, which has enabled dealing with a wide spectrum of human needs effectively. These needs vary from simple day-to-day needs like paying electricity bills, booking train tickets, etc., to complicated needs like sharing and power grids for power generation. These technologies have taken human life into much higher levels of sophistication and ease. However, in the middle of this phenomenon, the rise and growth of a parallel technology is startling that of compromising security, thereby resulting in different effects detrimental to the use of technology. This includes attacks on information, such as thieving of personal information, hacking, and outage of services. The aim of the work leading to this paper is to address the problem of DoS and DDoS detection at the target end using machine learning techniques. With emphasis on hardware implementation, the paper provides the following key contributions. They are (i) the systems approach for DoS and DDoS detection at target using a Naive Bayes classifier for TCP, with a design engineered for real time use and implementation. (ii) Light weight detection algorithm, with a note on processing latency and reaction time.

Login to Download PDF

Please login with your Paper ID and password to access the full PDF.

🔑 Login to Download