Recently, mobile devices have become paramount in online video streaming. Adaptive bitrate (ABR) algorithms of players responsible for selecting the quality of the videos face critical challenges in providing a high Quality of Experience (QoE) for end users. One open issue is how to ensure the optimal experience for heterogeneous devices in the context of extreme variation of mobile broadband networks. Additionally, end users may have different priorities on video quality and data usage (i.e., the amount of data downloaded to the devices through the mobile networks). A generic mechanism for players that enables specification of various policies to meet end users’ needs is still missing. In this paper, we propose a weighted sum model, namely WISH, that yields high QoE of the video and allows end users to express their preferences among different parameters (i.e., data usage, stall events, and video quality) of video streaming. WISH has been implemented into ExoPlayer, a popular player used in many mobile applications. The experimental results show that WISH improves the QoE by up to 17.6% while saving 36.4% of data usage compared to state-of-the-art ABR algorithms and provides dynamic adaptation to end users’ requirements.
WISH: User-centric Bitrate Adaptation for HTTP Adaptive Streaming on Mobile Devices
1. WISH: User-centric Bitrate Adaptation for HTTP
Adaptive Streaming on Mobile Devices
IEEE MMSP 2021
06.-08.10.2021 | Tampere, Finland
Minh Nguyen, Ekrem Cetinkaya, Hermann Hellwagner, Christian Timmerer
Christian Doppler laboratory ATHENA | Alpen-Adria University | Austria
minh.nguyen@aau.at | https://athena.itec.aau.at/
1
8. WISH: User-centric Bitrate Adaptation
8
● Throughput (data) cost of bitrate is a linearly increasing function
● Buffer cost increases when the download time increases and/or the
buffer level decreases.
9. WISH: User-centric Bitrate Adaptation
9
● Quality cost comprises two sub-penalties
○ Distortion penalty: When a representation is lower than the highest-
bitrate representation
○ Instability penalty: When that representation is different from the
average quality of recent segments
Distortion penalty Instability penalty
10. WISH: User-centric Bitrate Adaptation
10
● The overall cost of each representation is the weighted sum of
Throughput cost, Buffer cost, and Quality cost
● Selected representation has the lowest overall cost
11. Weights Determination
11
● Consider C(i) as the function of bitrate
● Weights are determined by making the maximum bitrate own the lowest
cost (i.e., the derivative of ) at particular conditions:
12. Weights Determination
12
● Without loss of generality
● For relaxation, we set Throughput cost = Buffer cost at the max bitrate
● Finally, the weights are defined as
14. Evaluation and Discussion
14
● Experimental setup
○ HAS testbed
○ 5-min test sequences with different SI and TI
○ Bitrate ladder: {107, 240, 346, 715, 1347, 2426, 4121} kbps
○ Codec: H265/HEVC
Apache Server
(Ubuntu)
Mobile Phone
(ExoPlayer)
4G network
(tc)
HAS testbed Test sequences
15. Evaluation and Discussion
15
● Results: Comparison with state-of-the-art approaches
○ WISH achieves the highest QoE scores for all test sequences
○ WISH’s QoE scores: from 3.46 (GamePlay) to 3.71 (ToS2)
○ Compared methods: < 3.40
⇒ QoE score: +17.6%
ITU-T QoE Score
16. Evaluation and Discussion
16
● Results: Comparison with state-of-the-art approaches
○ WISH has the fewest number of stalls with at most one stall
○ WISH: < 0.5 stalls with < 1.2s length
○ BBA-0 and SQUAD: > 2 stalls with average duration 15s to 30s
Number of stalls and stall duration
17. Evaluation and Discussion
17
● Results: Comparison with state-of-the-art approaches
○ WISH downloads the least bitrate
○ WISH: 2053 kbps ⇒ save 7.1% to 36.4% data usage
Average bitrate
18. Evaluation and Discussion
18
● Results: Comparison with state-of-the-art approaches
○ WISH and ExoPlayer: high video instability and # of switches
○ BBA-0 and SQUAD: the fewest # of switches and small instability
Video Instability
19. Evaluation and Discussion
19
● Results: WISH’s performance with different settings
○ Service providers meet their needs of data usage by varying the safe
threshold ξ
○ Higher ξ ⇒ smaller γ ⇒ less priority to high bitrate
○ Higher ξ ⇒ lower bitrate, less video instability, fewer switches and stalls
WISH’s performance with different ξ values
21. Conclusions
21
● WISH: a weighted sum model to provide high QoE for mobile devices and
to meet end users’ requirements
● Taking into account throughput cost, buffer cost, and quality cost of each
quality level
● A mathematical solution to choose those weights
● WISH needs the lowest data usage while keeping the highest QoE scores
● In the future, integrate the retransmission technique H2BR [1] to improve
the video stability of WISH
[1] Nguyen, M., Timmerer, C. and Hellwagner, H., 2020, June. H2BR: an HTTP/2-based retransmission technique to
improve the QoE of adaptive video streaming. In Proceedings of the 25th ACM Workshop on Packet Video (pp. 1-7).