State Estimation for Singularly Perturbed Semi-Markov Jump Coupled Neural Networks Under WTODP: A Privacy-Preserving Mechanism

This article studies the state estimation problem for singularly perturbed semi-Markov jump coupled neural networks. To save bandwidth resources and reduce communication frequency, a weighted try-once-discard protocol is recommended. The protocol prioritizes the transmission of node information with the largest weighted error, thereby enhancing the efficiency of information exchange within the network. At the same time, in the process of information interaction, a new encryption-decryption scheme is designed to effectively prevent the leakage of sensitive information while ensuring ideal estimation performance. For the singular perturbation parameter in the coupled neural network, this article uses the matrix contract transformation method to strictly realize the separation of the two dynamics of the system, thereby ensuring that the established mean square exponential stability criterion can be applied to the augmented system. Lastly, numerical simulations are employed to confirm the validity of the provided state estimator.