Adaptive control architectures for mitigating sensor attacks in cyber-physical systems

The accuracy of sensor measurements is critical to the design of high performance control systems since sensor uncertainties can significantly deteriorate achievable closed-loop dynamical system performance. Sensor uncertainty can arise due to low sensor quality, sensor failure, or detrimental environmental conditions. For example, relatively cheap sensor suites are used for low-cost, small-scale unmanned vehicle applications that can result in inaccurate sensor measurements. Alternatively, sensor measurements can also be corrupted by malicious attacks if dynamical systems are controlled through large-scale, multilayered communication networks as is the case in cyber-physical systems. This paper presents several adaptive control architectures for stabilization of linear dynamical systems in the presence of sensor uncertainty and sensor attacks. Specifically, we propose new and novel adaptive controllers for state-independent and state-dependent sensor uncertainties. In particular, we show that the proposed controllers guarantee asymptotic stability of the closed-loop dynamical system when the sensor uncertainties are time-invariant and uniform ultimate boundedness when the uncertainties are time-varying. We further discuss the practicality of the proposed approaches and provide a numerical example to illustrate the efficacy of the proposed adaptive control architectures.