Multi-Model Resilient Observer under False Data Injection Attacks
This addresses security vulnerabilities in cyber-physical systems, such as power grids, to prevent catastrophic failures from stealthy attacks, representing an incremental improvement in observer design.
The paper tackled the problem of false data injection attacks in cyber-physical systems by developing a resilient observer that uses auxiliary models to refine state estimation, and numerical experiments on an IEEE 14-bus system showed it successfully recovered true states.
In this paper, we present the concept of boosting the resiliency of optimization-based observers for cyber-physical systems (CPS) using auxiliary sources of information. Due to the tight coupling of physics, communication and computation, a malicious agent can exploit multiple inherent vulnerabilities in order to inject stealthy signals into the measurement process. The problem setting considers the scenario in which an attacker strategically corrupts portions of the data in order to force wrong state estimates which could have catastrophic consequences. The goal of the proposed observer is to compute the true states in-spite of the adversarial corruption. In the formulation, we use a measurement prior distribution generated by the auxiliary model to refine the feasible region of a traditional compressive sensing-based regression problem. A constrained optimization-based observer is developed using l1-minimization scheme. Numerical experiments show that the solution of the resulting problem recovers the true states of the system. The developed algorithm is evaluated through a numerical simulation example of the IEEE 14-bus system.