Secure State Estimation for a Class of Nonlinear Systems Over Sensor Networks With Sensor Resolution: Tackling Replay Attacks

This paper addresses the problem of distributed state estimation for nonlinear systems over sensor networks that are subject to replay attacks. Sensor resolution, recognized as a crucial indicator of measurement accuracy and its influence on estimation performance, is taken into account to reflect practical engineering scenarios. An upper-bounding technique is employed to handle the uncertainty introduced by sensor resolution. Replay attacks, executed by adversaries on communication channels between sensor nodes, are characterized by the replacement of current innovations with previously recorded innovations. The dynamic behavior of these replay attacks is described by two factors: one dependent on a stochastic variable and the other on a time-varying parameter. The objective of this study is to design distributed state estimators to ensure that the estimation error dynamics are exponentially ultimately bounded in the mean-square sense while the desired security level is maintained. Furthermore, the gain parameters of the estimators are determined by solving specific matrix inequalities. Finally, the feasibility and effectiveness of the proposed estimation approach are validated through numerical simulations.