Diagnosing network-wide P2P live streaming inefficiencies
Bibliographic record
Abstract
Large-scale live peer-to-peer (P2P) streaming applications have been successfully deployed in today's Internet. While they can accommodate hundreds of thousands of users simultaneously with hundreds of channels of programming, there still commonly exist channels and times where and when the streaming quality is unsatisfactory. In this paper, based on more than two terabytes and one year worth of live traces from UUSee, a large-scale commercial P2P live streaming system, we show an in-depth network-wide diagnosis of streaming inefficiencies, commonly present in typical mesh-based P2P live streaming systems. As the first highlight of our work, we identify an evolutionary pattern of low streaming quality in the system, and the distribution of streaming inefficiencies across various streaming channels and in different geographical regions. We then carry out an extensive investigation to explore the causes to such streaming inefficiencies over different times and across different channels/regions at specific times, by investigating the impact of factors such as the number of peers, peer upload bandwidth, inter-peer bandwidth availability, server bandwidth consumption, and many more. The original discoveries we have brought forward include the two-sided effects of peer population on the streaming quality in a streaming channel, the significant impact of inter-peer bandwidth bottlenecks at peak times, and the inefficient utilization of server capacities across concurrent channels. Based on these insights, we identify problems within the existing P2P live streaming design and discuss a number of suggestions to improve real-world streaming protocols operating at a large scale.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.017 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".