ghostnwa / detecting-ip-spoofing-attacks-using-deep-learning--ml Goto Github PK
View Code? Open in Web Editor NEWThis project forked from shubhagrwl/detecting-ip-spoofing-attacks-using-deep-learning--ml
Intrusion detection system is one of the most important parts of network security in competing against illegitimate network access. Intruder uses hijacking technique like host file hijack or IP spoofing, which is IP address forgery. There are different ways by which intrusion can take place, such as fake emailing and denial of service attack. In such kind of intrusion it is difficult to find the identity of the sender. Because of this, the system or the main server can crash from the overloading requests coming from an unauthorized source. Therefore, the system becomes unresponsive to the authorized requests. In the IP address spoofing, the attacker’s mimics as the legitimate user and access the network authorization. Further, it modifies the packet headers of the authorized users IP address. To mitigate this issue, in this thesis work, a new intrusion detection method has been introduced for preventing IP spoofing attack. The properties of nodes are optimized by using artificial bee colony optimization (ABC). Furthermore, the neural network model is trained using the optimized properties to predict the output in terms Packet delivery ratio (PDR), throughput and energy consumption.