A Hybrid Approach to Protect Server from IP Spoofing Attack

The rapid expansion of the internet has resulted in distributed denial of service cyberattacks being one of the most critical problems in a data centre environment where numerous servers are placed. A Distributed Denial of Service attack especially IP spoofing, carried out by countless specialists, creates a significant number of packets, which can quickly exhaust the processing and communication capabilities of a target in a short amount of time. Defending against such attack require a number of actions, including detection, classification, and traceback in order to mitigate the problem. In this research, we present a hybrid strategy for the identification and prevention of distributed denial of service (DDoS) attacks based on IP spoofing. Our suggested detection system makes use of both distance calculation and machine learning-based detection methods individually, but in an integrated manner, and it integrates the results of the two indicators to upgrade the system identification precision by combining their results. Our proposed hybrid system is tested against the CAIDA 2007 dataset in order to determine its detection performance. As a result, DDoS attack based on IP spoofing on the internal network is detected and spoofed IPs are found out and analyzed for the protection of server in order to forestall unapproved access and to maintain security of information handled by server. Also, SNORT based prevention technique is used so as to protect the server from attack. We conclude that the suggested hybrid system outperforms the non-hybrid detection systems in terms of detection performance and helps in preventing the attack by finding out the spoofed IP packet causing the attack thus protecting the system from attack.