Privacy and Security Control Approach for DDoS Attacks in Cyber Physical Systems using Deep Learning

Attacks that are classified as a distributed denial of service, or DDoS, are becoming an increasingly serious concern for organizations and government institutions. They hurt the economy that are based on the internet, restrict access to information and services, and undermine the reputation of companies. Impersonating real users allows attackers to launch application layer DDoS assaults that are difficult to detect due to the fact that they cannot be readily identified. In this study, we confront novel application-layer DDoS threats by examining the features of incoming packets. These characteristics include the dimensions of HTTP frame data packet, the amount of Internet Protocol (IP) addresses sent, continuous mapper of ports, and the amount of IP addresses using proxy IP. In addition, we look at the amount of IP addresses using proxy IP. The challenge of multi - layer security management for protecting complex cyber-physical systems from denial-of-service assaults is investigated in this work. On the foundation of piecewise Lyapunov functions, a proposal has been made for a distributed security controller that includes control nodes that are linked to one another. When trying to accomplish security synchronisation, it is necessary to explicitly determine the boundaries of attack frequency as well as attack duration. In the meantime, an approach for the event-based updating of information has been suggested as a defense against the disruptive effects of DoS assaults on data transmission. In conclusion, a simulations scenario is shown as a means of demonstrating the usefulness of the suggested strategy.