Latency Considerations in the Context of Traffic and Street Surveillance Using Adaptive Streaming
摘要
The Adaptive Streaming Framework (ASF) is a novel approach designed to optimize data transmission in traffic and street surveillance systems, a critical component of modern smart cities. By leveraging sensor virtualization and advanced wireless telecommunication networks, including 5G, ASF selectively streams only relevant data, significantly reducing network load, latency, and energy consumption while enhancing privacy. This paper presents an experimental setup involving high-dimensional sensors, embedded devices, and a 5G network to evaluate the impact of frame rates, exposure times, and network timing on system latency. Results demonstrate that higher frame rates and shorter exposure times are essential for achieving low latencies, critical for real-time applications like accident detection and adaptive traffic control. While the 5G network performs efficiently, the primary bottleneck lies in camera hardware and its integration with operating systems.