Asynchronous Event-Triggered Impulsive Decentralized Control of Complex-Valued Multilayer Large-Scale Systems Under Deception Attacks

This article develops an asynchronous event-triggered impulsive decentralized control scheme for a class of complex-valued multilayer large-scale systems with time-varying coupling strengths and delays, under deception attacks where adversaries achieve their objectives by tampering with historical data. Unlike existing studies on event-triggered impulsive control where all node controllers are activated synchronously, each node in the proposed framework independently determines its triggering instants based on its own state. Initially, the interconnection effects are neglected, and node-specific Lyapunov functions are constructed and analyzed. Subsequently, graph-theoretic techniques, combined with the Razumikhin method, are employed to handle the cross-coupling terms and to derive several sufficient conditions. These conditions guarantee the p-th moment exponential stability of the closed-loop system while explicitly accounting for communication delays and attack probabilities. Finally, the theoretical results are applied to inertial neural networks, and numerical simulations demonstrate the effectiveness of the proposed control strategy.