ncmprtp
Differences
This shows you the differences between two versions of the page.
Both sides previous revisionPrevious revisionNext revision | Previous revision | ||
ncmprtp [2017-04-20 07:36] – manlio | ncmprtp [2022-07-25 15:54] (current) – [PROPOSED MODIFIED SCHEDULER] manlio | ||
---|---|---|---|
Line 6: | Line 6: | ||
===== Scheduler Implementation ===== | ===== Scheduler Implementation ===== | ||
- | The scheduler implementation available at [2] estimates the overall available bandwidth, and splits the multimedia flow accordingly. The source video bitrate is then set, in order to provide the maximum achievable video quality, according to the available bandwidth. Anyway, [2] does not provide a way to reduce the load fraction on lossy and/or high-delay sub-links, thus causing corrupted video at the receiver. As analysed in [3], the scheduler should be able to quickly adapt to bandwidth changes, to overcome packet skew due to different RTTs, and to choose an optimal subset of sub-links, in order to provide the desired Quality of Service (QoS) and Quality of Experience (QoE). | + | The scheduler implementation available at [2] estimates the overall available bandwidth, and splits the multimedia flow accordingly. The source video bitrate is then set, in order to provide the maximum achievable video quality, according to the available bandwidth. Anyway, [2] does not provide a way to reduce the load fraction on lossy and/or high-delay sub-links, thus causing corrupted video at the receiver. As analysed in [3], the scheduler should be able to quickly adapt to bandwidth changes, to overcome |
==== ORIGINAL SCHEDULER [2] ==== | ==== ORIGINAL SCHEDULER [2] ==== | ||
{{: | {{: | ||
- | Corruption on even a single link within the pool makes the whole multimedia stream unusable. | + | |
+ | //Corruption on even a single link within the pool makes the whole multimedia stream unusable.// | ||
==== PROPOSED MODIFIED SCHEDULER ==== | ==== PROPOSED MODIFIED SCHEDULER ==== | ||
- | In order to reduce the impact of links showing bad channel statistics, we modified the scheduler available at [2], in order to only use the links within the pool that can deliver the multimedia stream in a usable way. | + | In order to reduce the impact of links showing bad channel statistics, we modified the scheduler available at [2], in order to only use the links within the pool that can deliver the multimedia stream in a usable way. Read the current advancements in [4,5,6,7]. |
{{: | {{: | ||
- | Corruption on a single link does not make the whole stream unusable, because the former is excluded from the subset of the links in use. | + | //Corruption on a single link does not make the whole stream unusable, because the former is excluded from the subset of the links in use. |
+ | // | ||
Line 27: | Line 28: | ||
[1] An IETF draft is available at https:// | [1] An IETF draft is available at https:// | ||
- | [2] MP-RTP implementation | + | [2] MP-RTP implementation: |
[3] Singh V, Ahsan S, Ott J. //MPRTP: multipath considerations for real-time media//. Proceedings of the 4th ACM Multimedia Systems Conference 2013 Feb 28 (pp. 190-201). | [3] Singh V, Ahsan S, Ott J. //MPRTP: multipath considerations for real-time media//. Proceedings of the 4th ACM Multimedia Systems Conference 2013 Feb 28 (pp. 190-201). | ||
+ | [4] Bacco, Manlio, Pietro Cassarà, Alberto Gotta, and Massimo Puddu. "A Simulation Framework for QoE-Aware Real-Time Video Streaming in Multipath Scenarios." | ||
+ | |||
+ | [5] Bacco, Manlio, Pietro Cassarà, Alberto Gotta, and Vincenzo Pellegrini. " | ||
+ | |||
+ | [6] Bacco, Manlio, Pietro Cassarà, and Alberto Gotta. " | ||
+ | |||
+ | [7] Machumilane, |
ncmprtp.1492673766.txt.gz · Last modified: 2017-04-20 07:36 by manlio