Explora-VR

Explora-VR implements edge-assisted content prefetching for tile-based 360-degree video streaming to reduce latency and prevent playback interruptions by prefetching tiles predicted by the client's rate adaptation heuristic and storing them on cache-enabled edge servers.


Key Features:

  • Dynamic Content Prefetching: Uses the client's rate adaptation heuristic to predict and prefetch tiles needed in upcoming segments to reduce playout freezes.
  • Incremental Buffer Building: Constructs a dynamic collective buffer for active streaming sessions based on estimated tile popularity per segment.
  • High Cache Efficiency: Serves over 98% of client requests from cache-enabled edge servers.
  • Performance Improvement: Increases perceived throughput by 2.5× versus conventional client-server setups and 1.4× versus least recently used (LRU) caching policies.
  • Quality Enhancement: Enables streaming of higher-quality video with freeze-free playback while reducing network traffic to the content server.

Scientific Applications:

  • Tile-based 360-degree video streaming: Reduces latency and prevents playback interruptions through predictive tile prefetching.
  • Edge-assisted caching for immersive video: Improves cache hit rates and offloads traffic from origin content servers.
  • Evaluation with head movement traces: Validates throughput and cache performance using head movement traces from 48 users across three 360-degree videos.

Methodology:

Prefetches tiles using the client's rate adaptation heuristic; estimates tile popularity per segment to incrementally build a collective buffer on edge servers; evaluation used head movement traces from 48 users across three 360-degree videos and comparisons to client-server and LRU caching.

Topics

Details

License:
Other
Cost:
Free of charge (with restrictions)
Tool Type:
workflow
Programming Languages:
Python, Shell
Added:
6/25/2022
Last Updated:
11/24/2024

Operations

Publications

Ordonez-Ante L, van der Hooft J, Wauters T, Van Seghbroeck G, Volckaert B, De Turck F. . J Netw Syst Manage. 2022;30(3).

PMCID: PMC8920064

Links