On the Integration of Multiagent and Reconfigurable Intelligent Surface-Assisted Networks in Smart Environments

This article presents an integrated framework for resilient wearable health monitoring in dynamic indoor environments by combining multi-agent control, Reconfigurable Intelligent Surfaces (RIS), and Voronoi tessellation. The framework partitions indoor space into Voronoi cells, updates relay placement via a Genetic Algorithm (GA), and triggers event-driven RIS reconfiguration to sustain links under mobility and blockage. In a Light Detection and Ranging (LiDAR)-scanned multi-room deployment, the proposed loop achieves a normalized Link Quality Indicator (LQI) of 0.91 (range [0,1]), reduces end-to-end latency to 12.5 ms, and attains a power-per-throughput of 0.31 mW/Mbps, with a power efficiency (normalized throughput-per-watt) of 0.893 (1.0 = best baseline). Against static, mobile-only, and RIS-only baselines, the composite performance score improves by +26.4, +15.7, and +11.9 percentage points (pp), respectively. Under denser conditions, the 95th-percentile latency remains below 16.9ms with less than a 3.8 pp reduction in coverage, suggesting limited degradation under higher density. A concise living-lab pilot protocol is included to guide real-world validation.