ESP32-Based Edge Computing Node for Real-Time Sound Pressure Level Monitoring in Smart Mosque Cyber-Physical Systems

Abstract

Audio distribution systems in mosques often fail to maintain consistent sound pressure levels (SPL) across zones, while conventional public address (PA) setups still rely on subjective, manual gain adjustments without quantitative monitoring. This paper presents an ESP32‑based edge computing node for real‑time SPL monitoring in smart mosque audio systems, designed as an accurate, low‑cost sensing layer within a three‑layer edge–fog–cloud cyber‑physical architecture. The node integrates an ESP32 microcontroller, an INMP441 digital MEMS microphone, and MQTT telemetry; it computes A‑weighted SPL locally using RMS over a 128 ms window and publishes results to a local broker at 1 Hz. Experimental evaluation covers SPL accuracy against a calibrated Class 2 sound level meter across seven representative test scenarios, as well as end‑to‑end telemetry latency and MQTT reliability in a laboratory environment. For 21 node–SLM measurement pairs, the node achieves a mean absolute error of 0.81 dB and an RMSE of 0.83 dB, with 90.5% of measurements within ±1 dB and all within ±2 dB of the reference, and a coefficient of determination close to 0.997. Telemetry tests show an average end‑to‑end latency of approximately 170.8 ms and a packet loss rate between 0% and 0.28% over a 60‑minute test at 1 Hz, with small inter‑arrival jitter around the nominal interval. These results indicate that the proposed SPL edge node provides sufficiently accurate and timely measurements for continuous monitoring and supervisory‑level control in mosque audio systems, and can serve as a practical sensing building block for future smart mosque implementations with distributed, zone‑aware audio management