Lizard cuts video delays and improves analysis when bandwidth drops
Lizard: Bandwidth-Adaptive Real-Time Video Analytics through Content-Aware Packet Discarding at Last-Mile Edge Routers
Networking and Internet ArchitectureDistributed, Parallel, and Cluster Computing
Summary
Video analytics running near users can slow down a lot if the internet connection gets worse, causing delays and less accurate results. The authors introduce Lizard, a system that smartly throws away less important parts of a video to keep things moving smoothly. It figures out which pieces are less useful for analysis and drops their packets first when bandwidth is tight. This approach reduces delays and even makes the analysis more accurate compared to other methods.
What this means in practice
- •For video streaming engineers: Improve real-time video analytics performance on networks with fluctuating last-mile bandwidth by dropping low-priority video data dynamically.
- •For network operators: Manage bandwidth congestion at edge routers by selectively discarding unimportant video packets to reduce latency without severely impacting analysis accuracy.
Authors
Shan Yu, Yu Chen, Yifan Qiao, Sheng Zhang, Ravi Netravali, Harry Xu
Abstract
The timeliness and accuracy of edge-based video analytics can be hindered by drastic reductions in available bandwidth (ABW) at last-mile edge routers, causing prolonged queuing delays. This work proposes Lizard, a system that leverages video-content-aware packet discarding to mitigate the negative effects of drastic ABW degradation that may frequently occur at a last-mile edge router by judiciously discarding packets that contain frame blocks less important to the analytics at the destination. To achieve this, we first devise a frame-block-aware RTP header extension to effectively decouple packet dependencies to encode frame blocks. Second, Lizard uses a priority-based feedback mechanism that dynamically evaluates packet priorities based on relative accuracy impacts. Third, we develop an adaptive phase-transition-based packet discarding strategy at the router to discard packets that represent unimportant blocks. Our evaluation of Lizard shows improvements over existing methods are substantial: 53.2% reduction in latency and 27.1% increase in analysis accuracy.