Distributed Finite-Time Filtering of Switched Positive Systems with Communication Delays and Deception Attacks

This article investigates the design of secure finite-time distributed filters, for linear discrete-time switched positive systems in the presence of deception attacks and communication delays. Consider a set of sensors with communication interaction for measuring the output information of switched positive systems. In an interactive sensor network, in the event of a random failure in the interactive communication topology, sensor network based on a given interactive communication topology can share its measurement values with its neighboring nodes. In addition, considering the impact of deception attacks on the aggregated measurement information transmitted through communication networks, the aggregated measurement values of sensors will be compromised. Based on the above issues, the following solutions are proposed. Firstly, a sensor transmission model is proposed based on the effects of random communication link failures, deception attacks, and communication delays. Secondly, the network communication delay is considered Markov process related, and the filters is modeled as a Markov jump system using an augmented matrix. Thirdly, considering the target system switching based on the ADT rule, a sufficient condition for the finite time performance of distributed residual systems is obtained by constructing linear multiple co-positive Lyapunov functional, and a scheme is provided for the design of distributed filters for switched positive systems. Finally, explained the feasibility of the system implementation scheme was verified through numerical simulation.