Clustering‐based quality selection heuristics for HTTP adaptive streaming over cache networks
Bibliographic record
Abstract
Summary HyperText Transfer Protocol (HTTP) Adaptive Streaming (HAS) has become the de facto standard video‐streaming technology. The benefits of HAS are manifold: reliable transmission of video data avoiding artifacts caused by packet loss, easy fire wall, and Network Address Translation (NAT) traversal and the seamless reuse of existing HTTP caching infrastructure. However, introducing transparent, intermediary caching nodes on the delivery path can impact the Quality of Experience (QoE) perceived by the end user. In cache‐assisted HAS, segments can be served from different origins based on the content of the caches, causing highly fluctuating throughput and Round‐Trip Time (RTT) measurements, negatively impacting the stability and optimality of the quality decisions due to incorrect throughput estimations. In this paper, we propose heuristics that are able to use information on the streaming origin and intermediary cache contents to optimize the quality selection process. Using more accurate per origin throughput measurements, buffer starvations can be avoided. Moreover, including the cache state information in the decision process can positively impact the streaming quality. Furthermore, approximation techniques based on unsupervised incremental clustering are proposed to detect the streaming origin in absence of an external information channel. Similarly, a cache probability‐based heuristic is proposed to predict the content of the expected delivery location when this information is not transferred. With perfect information, the proposed heuristics improve the QoE with 0.52 on a scale between 1 and 5, while the approximation techniques result in a performance gain between 0.04 and 0.36 for a dynamic scenario and a reduction of buffer starvations with a factor 3 to 7.
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 | 0.001 | 0.005 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 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".