Noncoherent Communication over Conductive Fabrics: Reliability and Energy Tradeoffs for Wearable Nodes

Wearable garments are evolving into distributed communication and electronic systems in which multiple sensing and actuation nodes exchange data and share power through the textile itself. Two parallel conductive fabric layers can serve this dual role, distributing a DC supply while carrying a narrowband communication channel. This paper investigates a system-level framework for such links, focusing on noncoherent BFSK implemented entirely on microcontrollers and peripheral modules without RF chipsets. The conductive fabric introduces slow, motion-driven fading caused by seams, stretch, and contact variation, as well as coexisting power-data paths. We model this environment with distance-dependent attenuation and slow Rician block fading, and derive closed-form reliability expressions for low-complexity coded modes, including repetition (3,1) and Hamming (7,4). A node-processing energy model is then combined with frame reliability to quantify the energy cost per successfully delivered bit. Monte-Carlo simulations validate the framework under representative operating conditions, with tones in the 100-300 kHz range and sampling at 1 MS/s. Measurements from a wearable prototype extract and confirm the assumed model statistics and PER trends. The model yields practical inter-node spacing, payload sizing, detector window, and coding methods, so developers can estimate packet-error rate and energy trade-offs before conducting fabric garment prototyping.