Modeling and analyzing the influence of chunk size variation on bitrate adaptation in DASH

Zhang, Tong; Ren, Fengyuan; Cheng, Wenxue; Luo, Xi; Shu, Ran; Liu, Jia

doi:10.1109/infocom.2017.8057057

Cited by 27 publications

(12 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This is because DDS detects most of the objects in Stream A whereas AWStream sends a single video stream and spends more time to transmit that stream through the network. To put it into perspective, popular video sites use 4-second to 8-second segments [81] (i.e., over 120 frames per segment), a range in which DDS's gains are considerable.…”

Section: End-to-end Improvementsmentioning

confidence: 99%

Server-Driven Video Streaming for Deep Learning Inference

Pervaiz

Yuan

et al. 2020

Proceedings of the Annual Conference of the ACM Special Interest Group on Data Communication on the Applications, Technologies,

135

View full text Add to dashboard Cite

Video streaming is crucial for AI applications that gather videos from sources to servers for inference by deep neural nets (DNNs). Unlike traditional video streaming that optimizes visual quality, this new type of video streaming permits aggressive compression/pruning of pixels not relevant to achieving high DNN inference accuracy. However, much of this potential is left unrealized, because current video streaming protocols are driven by the video source (camera) where the compute is rather limited. We advocate that the video streaming protocol should be driven by real-time feedback from the server-side DNN. Our insight is twofold: (1) server-side DNN has more context about the pixels that maximize its inference accuracy; and (2) the DNN's output contains rich information useful to guide video streaming. We present DDS (DNN-Driven Streaming), a concrete design of this approach. DDS continuously sends a low-quality video stream to the server; the server runs the DNN to determine where to re-send with higher quality to increase the inference accuracy. We find that compared to several recent baselines on multiple video genres and vision tasks, DDS maintains higher accuracy while reducing bandwidth usage by upto 59% or improves accuracy by upto 9% with no additional bandwidth usage.

show abstract

Section: End-to-end Improvementsmentioning

confidence: 99%

Server-Driven Video Streaming for Deep Learning Inference

Pervaiz

Yuan

et al. 2020

Proceedings of the Annual Conference of the ACM Special Interest Group on Data Communication on the Applications, Technologies,

135

View full text Add to dashboard Cite

show abstract

“…As a result, a myopic scheme oftentimes mechanically selects very high (low) levels for chunks with small (large) sizes -exactly the opposite to what is desirable for VBR videos ( §3.3). We illustrate this using two myopic schemes: (i) BBA-1 [14], a bufferbased scheme, and (ii) a rate-based scheme [28], referred to as RBA henceforth. Both schemes are myopic: when selecting the track for the next chunk, BBA-1 selects the highest track based on a chunk map, which defines the allowed chunk sizes as a range from the average chunk size of the lowest track to that of the highest track; RBA selects the highest track so that after downloading the corresponding chunk, the player buffer will still contain at least four chunks, where the downloading time of a chunk is obtained as its size divided by the estimated network bandwidth.…”

Section: Design Principlesmentioning

confidence: 99%

“…There is a large body of existing work on ABR streaming [9,11,12,13,14,18,27,28,34,35,56,77,78,79,80]. Rate-based (e.g., [28,34,81]) ABR schemes select bitrate simply based on network bandwidth estimates. Buffer-based schemes, e.g., BBA [14],…”

Section: Related Workmentioning

confidence: 99%

“…As we have seen in §3.6.8, such simplistic treatment can lead to poor performance. Of the large number of existing studies, only a few targeted VBR videos [14,27,28]; most others were designed for and evaluated in the context of CBR encodings. While several schemes in the latter category can be applied for VBR videos by explicitly incorporating the actual bitrate of each chunk, they were not designed ground up to account for VBR characteristics, nor had their performance been evaluated for that case.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Adaptive Bitrate Streaming Over Cellular Networks: Rate Adaptation and Data Savings Strategies

Qin¹

2021

Preprint

View full text Add to dashboard Cite

Adaptive bitrate streaming (ABR) has become the de facto technique for video streaming over the Internet. Despite a flurry of techniques, achieving high quality ABR streaming over cellular networks remains a tremendous challenge. First, the design of an ABR scheme needs to balance conflicting Quality of Experience (QoE) metrics such as video quality, quality changes, stalls and startup performance, which is even harder under highly dynamic bandwidth in cellular network. Second, streaming providers have been moving towards using Variable Bitrate (VBR) encodings for the video content, which introduces new challenges for ABR streaming, whose nature and implications are little understood. Third, mobile video streaming consumes a lot of data. Although many video and network providers currently offer data saving options, the existing practices are suboptimal in QoE and resource usage. Last, when the audio and video tracks are stored separately, video and audio rate adaptation needs to be dynamically coordinated to achieve good overall streaming experience, which presents interesting challenges while, somewhat surprisingly, has received little attention by the research community. In this dissertation, we tackle each of the above four challenges.Firstly, we design a framework called PIA (PID-control based ABR streaming) that strategically leverages PID control concepts and novel approaches to account for the various requirements of ABR streaming. The evaluation results demonstrate that

show abstract

“…Dividing a video having a variable bitrate (VBR) into multiple segment results in segments with different sizes [21]. Owing to the segment size differences, the downloading time for a segment is altered.…”

Section: Qoe-based Block Quality Adaptationmentioning

confidence: 99%

Dynamic Block-Based HTTP Adaptive Streaming Scheme in Multiple Wireless Networks

2020

View full text Add to dashboard Cite

Multipath-based hypertext transfer protocol (HTTP) adaptive streaming (MPHAS) has been studied due to the increasing use of mobile devices equipped with multiple network interfaces and the prevalence of high-quality video streaming services. However, the existing MPHAS schemes undergo a degraded quality of experience (QoE) in multiple wireless networks. The segment request-based MPHAS (SR-MPHAS) encounters frequent quality switching, and the block request-based MPHAS (BR-MPHAS) causes results in playback stalling and inefficient bandwidth utilization, according to slow responsiveness. In this paper, we propose a dynamic block-based MPHAS (DBAS) to improve the QoE by overcoming the limitations of the existing schemes in multiple wireless networks. We propose buffer-based block length adaptation and QoE-based block quality adaptation to reduce the frequency of quality switching and to improve the responsiveness to network changes. Moreover, we propose a partial request-based block scheduler to mitigate segment reordering. The experimental results show that the DBAS improves the QoE by reducing the quality switching frequency and rapidly adapting to abrupt changes in the bandwidth.INDEX TERMS HTTP adaptive streaming, multipath-based transmission, multiple wireless networks, quality of experience.

show abstract

Modeling and analyzing the influence of chunk size variation on bitrate adaptation in DASH

Cited by 27 publications

References 13 publications

Server-Driven Video Streaming for Deep Learning Inference

Server-Driven Video Streaming for Deep Learning Inference

Adaptive Bitrate Streaming Over Cellular Networks: Rate Adaptation and Data Savings Strategies

Dynamic Block-Based HTTP Adaptive Streaming Scheme in Multiple Wireless Networks

Contact Info

Product

Resources

About