A distributed event-triggered framework for coordinating athlete autonomic states via wearable sensor networks

Interpersonal autonomic physiology (IAP) research has established that physiological synchrony (PS), the temporal interdependence of autonomic nervous system (ANS) activity across individuals, is an established phenomenon in team-based settings. However, PS in athlete cohorts has been studied exclusively as a passive observational variable; no framework actively regulates autonomic coherence under the communication constraints of wireless body area networks. This study integrates IAP and distributed control theory by formulating team-level autonomic coherence as a leaderless multi-agent consensus problem.

Each wearable unit was modeled as a two-dimensional linear agent whose state encodes cardiovascular autonomic tone (parasympathetic index: HRV-RMSSD) and sympathetic arousal load (EDA skin-conductance level, SCL), consistent with the Banister Fitness-Fatigue Model. Initial model states were derived from ECG/EDA records of WESAD subjects S2-S7, used only to provide heterogeneous simulation initial conditions (correlations were directional, non-significant sanity checks: r = -0.71, p = 0.114 for RMSSD; r = 0.68, p = 0.137 for SCL; N = 6). Three engineering challenges were addressed: (i) BLE communication delays h _c ∈ [15, 150] ms; (ii) neuromuscular input delays h _i ∈ [150, 300] ms; and (iii) switching directed topologies from dynamic team formations. A distributed dynamic event-triggered mechanism (DETM) drove all agents toward consensus while transmitting data only upon detection of meaningful state deviation. Stability was established via a dual-integral Lyapunov-Krasovskii functional, with LMI feasibility verified by CVXPY (ε = 0.087). Simulations ran for 300 s under three delay scenarios with five initial-condition perturbation folds; battery life was estimated from a four-component nRF52840 energy model.

The proposed DETM reduced BLE transmissions by 38.38% in the nominal case while achieving convergence of the normalized model states within 4-7 s. A Zeno-free guarantee (T _min = 0.06 s) and a theoretical battery-life estimate of about 36 days were obtained. The ablation analysis attributed the largest communication saving to the dynamic internal variable ρ _i (t) and the largest convergence-time improvement to the dual-delay structure. Five initial-condition perturbation folds indicated numerical robustness to the starting states.

These findings support control-theoretic and network-level feasibility only. Athlete studies are required before physiological benefits, safety, or real-world effectiveness can be inferred.