The coordination of decentralized multi-agent systems (MAS) operating under stochastic dynamics and time-varying network structures has become a critical challenge in modern control theory. This paper investigates a cyclic signal-initiated coordination framework for probabilistic decentralized systems subject to varying network configurations. The proposed approach integrates periodic event-triggered mechanisms with stochastic stability theory to ensure robust consensus and coordination in the presence of communication delays, packet losses, and structural topology variations. Unlike conventional continuous communication strategies, the cyclic signal-driven mechanism reduces communication burden by initiating information exchange only at discrete intervals determined by system states and probabilistic thresholds.

The framework incorporates adaptive control laws and stochastic modeling to capture uncertainties arising from environmental disturbances and network-induced imperfections. A unified theoretical model based on Markovian switching systems is developed to represent varying communication topologies, while Lyapunov-based methods are employed to establish mean-square stability and convergence properties. Furthermore, the proposed strategy addresses the limitations of asynchronous triggering and communication noise by integrating predictive estimation and state-dependent triggering conditions.

Comparative analysis demonstrates that the proposed method outperforms existing event-triggered and time-triggered coordination strategies in terms of communication efficiency, robustness, and convergence speed. Simulation-based evaluations indicate that the system maintains stable coordination under dynamic topology switching and stochastic disturbances, achieving significant reductions in communication overhead without compromising performance.

This study contributes to the advancement of decentralized control by providing a comprehensive framework that bridges cyclic signal-driven coordination with probabilistic system modeling. The results have significant implications for applications in autonomous systems, distributed robotics, and networked cyber-physical infrastructures, where efficient and resilient coordination is essential.