This paper proposes security techniques for detecting and counteracting attacks in Networked Control Systems (NCSs). A partially homomorphic based encryption technique, namely the Paillier scheme, is used to protect the confidentiality of the transmitted data over the network against eavesdropping attacks. An integrated secure and private observer-based controller with guaranteed stability together with a secure and private observer-based detection system has been developed. We start with an encrypted observer. Then, an observer-based encrypted controller is added to stabilize the system. Besides, an encrypted observer-based detection algorithm has been used to detect false data injection attacks. Moreover, in our design, the computations on the observer, the controller, and the decision-maker operate on encrypted data while protecting against eavesdropping attacks. Using digital processors in cyber systems leads to quantized transmitted signals through the network and also to design the observer and controller operations in such a way that they work on quantized data efficiently. Limitations on the parameters of the quantizations and the encryption scheme are given to ensure the stability of the closed-loop system and guarantee reliable bounds on the closed-loop performance. Simulation results using Tennessee Eastman Process demonstrate the effectiveness of the proposed scheme.

本文提出了用于检测和抵制网络控制系统 (NCS) 中的攻击的安全技术。一种基于部分同态的加密技术，即 Paillier 方案，用于保护通过网络传输的数据的机密性，防止窃听攻击。已经开发了一种集成的安全和基于私人观察者的控制器，具有保证的稳定性以及安全和基于私人观察者的检测系统。我们从一个加密的观察者开始。然后，添加一个基于观察者的加密控制器来稳定系统。此外，一种基于加密观察者的检测算法已被用于检测虚假数据注入攻击。此外，在我们的设计中，观察者、控制器和决策者的计算对加密数据进行操作，同时防止窃听攻击。在网络系统中使用数字处理器会导致通过网络量化传输信号，并且还设计观察器和控制器操作，以便它们有效地处理量化数据。给出了量化参数和加密方案的限制，以确保闭环系统的稳定性并保证闭环性能的可靠界限。使用田纳西伊士曼过程的仿真结果证明了所提出方案的有效性。