State Estimation with Model-Mismatch-Based Secrecy against Eavesdroppers

This study takes into consideration remote state estimation, where the state of a system is to be shared with a number of authorized users for any purpose (e.g., tracking, control), in the presence of eavesdroppers. We propose a novel control-theoretic secrecy mechanism to securely transmit the state estimate among the authorized users in the system. Moreover, as there isn't any cryptographic mechanism applied to the shared information in the conventional sense, it is not possible for the eavesdroppers to understand that the state estimate is hidden. A use-case of the proposed secrecy mechanism for a target tracking example is also demonstrated.