Secure fusion estimation against FDI sensor attacks in cyber–physical systems

Abstract

This paper is concerned with the secure fusion estimation problem in cyber–physical systems (CPSs), where false data injection attack (FDIA) is able to falsify measurement signals transmitted over network. In this paper, we consider that adversaries are not able to attack all measurements, i.e., several measurements remain not being attacked. Under this condition, local augmented systems including FDIA signals are respectively constructed with un-attacked measurements. For each augmented system, we design the joint Kalman fusion estimator, under the linear minimum variance sense, to simultaneously estimate FDIA and system states, which is fulfilled by introducing a compensation factor. Finally, a simulation example is given to verify effectiveness and advantages of the secure fusion estimation method.