Distributed State Estimation for Nonlinear Dynamical Networks With Stochastic Topological Structures Subject to Random Deception Attacks and Bit-Rate Constraints

• Published in: IEEE transactions on systems, man, and cybernetics. Systems Vol. 55; no. 6; pp. 3976 - 3988
• Main Authors: Hu, Jun, Xu, Bing, Caballero-Aguila, Raquel, Jia, Chaoqing, Dong, Hongli
• Format: Journal Article
• Language: English
• Published: IEEE 01.06.2025
• Subjects: Bit rate; Bit-rate constraints; deception attacks; Decoding; Digital communication; Electronic mail; Network topology; nonlinear complex dynamical networks (NCDNS); optimized state estimation; State estimation; Stochastic processes; stochastic topological structures; Switches; Topology; Upper bound
• ISSN: 2168-2216; 2168-2232
• DOI: 10.1109/TSMC.2025.3547926

In this article, the issue of distributed optimized state estimation under bit-rate constraints (SEBRCs) is studied for nonlinear complex dynamical networks (NCDNs) with stochastic topological structures and deception attacks. The information of each node is transmitted to the remote estimator through shared digital communication networks. Taking the bandwidth-limited situation into account, the model of bit-rate constraints and the encoding-decoding strategy are employed to reflect the principles of resource allocation and data schedule. Moreover, two Bernoulli sequences are adopted to depict the topologies switched stochastically and the deception attacks occurred randomly. A novel optimized SEBRCs method for NCDNs is proposed such that, for both stochastic topological structures and deception attacks, the covariance upper bound of estimation error can be derived and the estimator parameter can be determined accordingly. It is worth mentioning that both the related effects caused by attack probability and bit-rate constraints onto estimation performance are clarified from the monotonicity analysis perspectives, where new analysis method is given. Besides, a sufficient criterion is given to guarantee the uniform mean-square boundedness of the obtained covariance upper bound. Finally, the applicability of the presented SEBRCs method is demonstrated via solving the indoor localization problem with multiple mobile robots.