ASTRA helps video cameras adjust stream quality for better smart city monitoring
ASTRA: Low-Overhead Runtime Architecture for STReam Adaptation in Video Analytics
Networking and Internet Architecture
Summary
Processing live video from many cameras at the highest quality can slow down important tasks because it uses too much computing power and network bandwidth. To solve this, the authors created ASTRA, a system that changes the video quality in real time based on how busy the network is and what the camera sees. ASTRA works with different adjustment methods and was tested using real street cameras and a simulated setup. The tests showed that ASTRA keeps the video analytics accurate and fast over 90% of the time while using only a small amount of extra computing power. This means ASTRA can help smart city systems run smoothly without needing expensive upgrades.
video analyticsreal-time processingcamera adaptationresolutionframe ratenetwork conditionsedge computingGPU utilizationsmart city applicationsstreaming optimization
Authors
Mahshid Ghasemi, Zoran Kostic, Javad Ghaderi, Gil Zussman
Abstract
Real-time video analytics is crucial for smart city applications and cloud-connected vehicle control. To improve analytics accuracy, it is desirable to process the video at the highest resolution and frame rate. However, due to limited resources, streaming and processing video at the highest resolution and frame rate from all cameras is not feasible and adversely affects the analytics latency. Intelligent adaptation of cameras' resolutions and frame rates based on network conditions and the video content is crucial in order to optimize the performance. In this paper, we present ASTRA, a low-overhead runtime architecture for online adaptation of live camera analytics at the edge. ASTRA can execute various online algorithms as a black box. We deployed ASTRA in the realistic NSF COSMOS testbed and uniquely assessed its real-time performance using COSMOS' street-level cameras. We further evaluated ASTRA with up to eight emulated cameras by streaming a comprehensive video dataset under real-world network conditions. We used ASTRA's architecture to evaluate the practical performance of several classes of adaptation algorithms, including theoretical and empirical methods. The results indicate that ASTRA can provide system reliability (i.e., the probability of meeting accuracy and latency requirements) of more than 90% while maintaining performance within a deviation of less than 10% from optimal offline performance with average GPU utilization overhead of around 2% per camera.