H∞H∞ state estimation for memristive neural networks under variance constraint: A binary encoding scheme with probabilistic bit flips

This paper addresses the state estimation problem for memristive neural networks subject to time‐delay under variance constraint. To enhance the robustness of transmitted data, a binary encoding scheme is adopted for data transmission over the binary symmetric channel. Due to channel noise, random bit errors might arise when the binary string is transmitted through the memoryless binary symmetric channel. To characterize these bit errors, a sequence of Bernoulli distributed random variables is employed. The main objective is to propose a new variance‐constrained state estimation strategy without utilizing the augmentation method, and sufficient conditions are obtained to ensure the described performance requirement and the estimation error variance constraints. Finally, the feasibility of the newly developed variance‐constrained state estimation method is illustrated by a simulation example.