Security Control of Denial-of-Service Attacks in Cyber-Physical Systems Based on Dynamic Feedback.

Abstract

With the integration and development of multiple disciplines, as well as a large amount of research on multilevel interconnection networks, wireless sensor networks have become a new access point and have attracted widespread attention worldwide. However, limited by the characteristics of the wireless sensor network itself, the energy problem has repeatedly restricted its function, which needs to be solved urgently. This paper mainly improves the Geographical and Energy-Aware Routing (GEAR) based on cyber-physical systems and energy-aware routing protocols, which reduces the resource consumption of network nodes, achieves network power balance, and further extends the system life cycle. Although the open sharing of the network layer makes CPS more efficient in actual use, it also increases the risks faced by CPS. Among these potential risks, cyberattacks have become the main security problem that CPS must solve with the greatest destructive power. Among them, Denial of Service (DoS) is the simplest and most destructive type of network attack. We can allow the system to reject normal requests for sending and receiving data using the network resources of the communication channel, thereby performing security control. From the perspective of control, this paper studies the security control of CPS in DOS attacks. Firstly, considering the diversity of CPS packet loss factors, assuming that there is an upper limit on the probability characteristics of the total number of packet loss development, the Markov process of inherent packet loss and the Markov process of DoS attack satisfying this constraint are analyzed, and the above two independent packet loss factors are described on this basis. Two independent packet loss factor descriptions are given in the following. In this paper, through the research of wireless sensor networks, it is applied to the cyber-physical system and the cyber-physical system is further improved.