DESIGN AND IMPLEMENTATION OF A LOW-LATENCY EDGE GATEWAY MIDDLEWARE FOR SPARKLINK-BASED V2I SYSTEMS

Abstract

Vehicle-to-Infrastructure (V2I) communication systems demand ultra-low latency, high throughput, and reliable data delivery for safety-critical traffic applications, such as dynamic water level assessment and vehicle clearance evaluation at entryway barriers. Traditional wireless communication technologies, including Wi-Fi and Bluetooth Low Energy (BLE), frequently suffer from substantial channel congestion, connection setup delays, and elevated packet loss under dense traffic or severe environmental conditions. To overcome these limitations, this paper proposes a lightweight, high-throughput edge gateway middleware specifically designed for SparkLink Low Energy (SLE) short-range communication systems. The proposed middleware features a custom ten-byte data frame structure incorporating Modbus CRC16 error checking, alongside a decoupled multi-threaded processing pipeline using thread-safe queues. The middleware successfully processed all 70,000+ frames without any frame loss. It further integrates asynchronous local caching via an SQLite database in Write-Ahead Logging (WAL) mode with real-time forwarding over Message Queuing Telemetry Transport (MQTT). Experimental evaluations under stress-testing conditions exceeding 70,000 frames demonstrate that the proposed edge gateway achieves an average frame parsing latency of 70.5 microseconds. These findings confirm the system's ability to support deterministic, real-time edge processing and safety warnings in dynamic V2I application environments.