Event-Based Distributed Set-Membership Estimation for Complex Networks: A Coding-Decoding Method

This paper is concerned with the event-based distributed set-membership estimation problem for complex networks subject to unknown-but-bounded noise. To improve the security and robustness of data transmission and reduce the communication burden, a coding-decoding communication mechanism and a discrete-time dynamic periodic event-triggered mechanism are introduced into the communication channels. To be specific, a coder converts the transmitted data into specific codewords, which are then decoded by the corresponding decoder. The discrete-time dynamic periodic event-triggered mechanism determines whether the decoded values are released at certain periodic sampling instants. Thereafter, distributed setmembership estimators are designed to determine ellipsoidal sets that contain the system states, where the proposed estimators are truly distributed with a desirable scalability. Simultaneously, by utilizing mathematical induction, ellipsoidal set theory, and recursive optimization algorithms, it is possible to derive sufficient conditions for the existence of distributed set-membership estimators. By solving optimization problems, the estimator gain matrices and corresponding parameters are calculated. In the end, an illustrative example is provided to exhibit the effectiveness of the proposed set-membership estimation method.