CS2P: Improving Video Bitrate Selection and Adaptation with Data-Driven Throughput Prediction
Open accessThe DOI matches its Crossref record; 38 of the 38 references it cites were found in a registry or an index.
Abstract
Bitrate adaptation is critical in ensuring good users’ quality-of-experience (QoE) in Internet video delivery system. Several efforts have argued that accurate throughput prediction can dramatically improve (1) initial bitrate selection for low startup delay and high initial resolution; (2) midstream bitrate adaptation for high QoE. However, prior ef- forts did not systematically quantify real-world throughput predictability or develop good prediction algorithms. To bridge this gap, this paper makes three key technical contributions: First, we analyze the throughput characteristics in a dataset with 20M+ sessions. We find: (a) Sessions sharing similar key features (e.g., ISP, region) present similar initial values and dynamical patterns; (b) There is a natural “stateful” dynamical behavior within a given session. Second, building on these insights, we develop CS2P, a better throughput prediction system. CS2P leverages data-driven approach to learn (a) clusters of similar sessions, (b) an initial throughput predictor, and (c) a Hidden-Markov-Model based midstream predictor modeling the stateful evolution of throughput. Third, we develop a prototype system and show by trace-driven simulation and real-world experiments that CS2P outperforms state-of-art by 40% and 50% median pre- diction error respectively for initial and midstream through- put and improves QoE by 14% over buffer-based adaptation algorithm.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“And deep learning further empowers more effective local analysis and prediction [130] at the edge node of industry instead of the cloud.”
From Deep learning for edge computing applications: A State-of-the-Art survey (2020) · cites this paper as [130]
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“It has been shown in recent work [432] that accurate throughput prediction can significantly improve the QoE for adaptive video streaming.”
From A comprehensive survey on machine learning for networking: Evolution, applications and research opportunities (2018) · cites this paper as [432]
The cited source says what the claim attributes to it.
In the citing paper
“It is also shown that CS2P achieves 3.2% improvement on overall QoE and 10.9% higher average bitrate over state-of-art Model Predictive Control (MPC) approach, which uses harmonic mean for throughput prediction.”
From A comprehensive survey on machine learning for networking: Evolution, applications and research opportunities (2018) · cites this paper as [432]
In the citing paper
“Data-driven approaches, in particular ML, have been employed mainly in two different ways: (1) to predict changes in network QoS, namely throughput, and trigger adaptation mechanism to reduce rebuffering time [432], and (2) to select appropriate adaptation action [102].”
From A comprehensive survey on machine learning for networking: Evolution, applications and research opportunities (2018) · cites this paper as [432]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Book Review: Pattern Recognition and Machine Learning
Nasrabadi, N. M. (2007) · Journal of Electronic Imaging · doi:10.1117/1.2819119 · Cited by 22,066
Pattern Recognition and Machine Learning
(2006) · doi:10.1007/978-0-387-45528-0 · Cited by 9,824
Scikit-learn: Machine Learning in Python
(2011) · Journal of Machine Learning Research · doi:10.5555/1953048.2078195 · Cited by 8,539
(2006) · Springer eBooks · Cited by 8,123
Modeling TCP throughput: a simple model and its empirical validation
Padhye, J., Firoiu, V., Towsley, D., & Kurose, J. (1998) · Proceedings of the ACM SIGCOMM '98 conference on Applications, technologies, architectures, and protocols for computer communication · doi:10.1145/285237.285291 · Cited by 2,106 · cited 1×
“Throughput prediction: Prior work either developed approximate analytical models of TCP throughput as a function of packet loss and delay [24, 38] or leveraged Time-series models (e.”
MethodVivaldi: a decentralized network coordinate system
Dabek, F., Cox, R., Kaashoek, F., & Morris, R. (2004) · Proceedings of the 2004 conference on Applications, technologies, architectures, and protocols for computer communications · doi:10.1145/1015467.1015471 · Cited by 1,154 · cited 1×
“Seen in a broader context, CS2P can be regarded as a middle ground between centralized video control platforms (e.g., C3 [23], CFA [29]) and decentralized player-based bi-trate adaptation (e.g., BB [27], FESTIVE [30], MPC [47]).”
BackgroundA Control-Theoretic Approach for Dynamic Adaptive Video Streaming over HTTP
Yin, X., Jindal, A., Sekar, V., & Sinopoli, B. (2015) · Proceedings of the 2015 ACM Conference on Special Interest Group on Data Communication · doi:10.1145/2785956.2787486 · Cited by 1,016 · cited 6×
“…adaptation: While it is possible to develop adaptation approaches without using through-put estimation (e.g., using only the playback buffer occupancy [27]), recent work [47] argues that throughput-aware bitrate adaptation can deliver a better QoE than pure buffer-occupancy based approaches.”
BackgroundMethodInfluentialSelf-citationA buffer-based approach to rate adaptation: evidence from a large video streaming service
Huang, T.-Y., Johari, R., McKeown, N., Trunnell, M., & Watson, M. (2014) · Proceedings of the 2014 ACM conference on SIGCOMM · doi:10.1145/2619239.2626296 · Cited by 1,004
Generalized Boosted Models: A guide to the gbm package
Ridgeway, G. K. (2006) · Cited by 814
Understanding the impact of video quality on user engagement
Dobrian, F., Sekar, V., Awan, A., Stoica, I., Joseph, D., Ganjam, A., Zhan, J., & Zhang, H. (2011) · Proceedings of the ACM SIGCOMM 2011 conference · doi:10.1145/2018436.2018478 · Cited by 673 · cited 2×
“Delivering good application-level video quality-of-experience (QoE) entails new metrics such as low buffering or smooth bitrate delivery [5, 22].”
BackgroundMethodSelf-citation
Research works printed but not linked: 3
- J. Jiang, V. Sekar, H. Milner, D. Shepherd, I. Stoica, and H. Zhang. CFA: A Practical Prediction System for Video QoE Optimization. In Proc. USENIX NSDI, 2016.
- H. V. Madhyastha, T. Isdal, M. Piatek, C. Dixon, T. Anderson, A. Krishnamurthy, and A. Venkataramani. iPlane: An Information Plane for Distributed Services. In Proc. USENIX OSDI, 2006.
- K. Murphy and M. Dunham. PMTK: Probabilistic Modeling Toolkit. In Proc. NIPS, 2008.
Cited by
Items 201–300 of 428
Chen, B.-R., Liu, Z., Song, J., Zeng, F., Zhu, Z., Bachu, S. P. K., & Hu, Y. (2022) · ACM International Conference on emerging Networking EXperiments and Technologies (CoNEXT) · doi:10.1145/3555050.3569139 · Cited by 10
Data-Driven QoE Analysis on Video Streaming in Mobile Networks
Wang, Q., Dai, H., Wang, H. H., & Wu, D. (2017) · IEEE International Symposium on Parallel and Distributed Processing with Applications (ISPA) · doi:10.1109/ispa/iucc.2017.00168 · Cited by 10
“The work of [11] improved the video bitrate adaptation based on data-driven QoE prediction.”
MethodKan, N., Liu, C., Zou, J., Li, C., & Xiong, H. (2019) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2019.8803007 · Cited by 10
“Note that the future throughput can be approximately predicted with the historical statistics [9].”
BackgroundYaqoob, A., & Muntean, G. (2024) · IEEE Transactions on Network and Service Management · doi:10.1109/tnsm.2024.3450014 · Cited by 9
“The adaptation algorithms attempt to optimize bandwidth utilization by selecting the highest available video bitrate, which is less than the available connection speed [47].”
MethodCratus: A Lightweight and Robust Approach for Mobile Live Streaming
Wang, B., Xu, M., Ren, F., Zhou, C., & Wu, J. (2021) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2020.3048826 · Cited by 9
“10, 11, 12, and 13.”
BackgroundMEC-based QoE Optimization for Adaptive Video Streaming via Satellite Backhaul
Cai, H., Jing, W., Wen, X., Lu, Z., & Wang, Z. (2021) · IEEE International Conference on Communications workshops · doi:10.1109/iccworkshops50388.2021.9473854 · Cited by 9
Proactive Energy-Aware Adaptive Video Streaming on Mobile Devices
Meng, J., Xu, Q., & Hu, Y. (2021) · USENIX Annual Technical Conference · Cited by 27
Xatu: Richer Neural Network Based Prediction for Video Streaming
Nam, Y. S., Gao, J., Bothra, C., Ghabashneh, E., Rao, S., Ribeiro, B., Zhan, J., & Zhang, H. (2021) · Proceedings of the ACM on Measurement and Analysis of Computing Systems · doi:10.1145/3491056 · Cited by 9
“CS2P [57] also observed that sessions sharing a combination of similar features such as ISP, CDN, user location, and timeof-day tend to have similar network characteristics.”
BackgroundMethodInfluentialViChaser: Chase Your Viewpoint for Live Video Streaming With Block-Oriented Super-Resolution
Chen, N., Zhang, S., Ma, Z., Chen, Y., Jin, Y., Wu, J., Qian, Z., Liang, Y., & Lu, S. (2023) · IEEE/ACM Transactions on Networking · doi:10.1109/tnet.2023.3286108 · Cited by 9
“Even thought several smoothing heuristics [8] are used to assist the throughput estimation, accurate throughput prediction remains a nonnegligible challenge.”
BackgroundVeritas: Answering Causal Queries from Video Streaming Traces
Bothra, C., Gao, J., Rao, S., & Ribeiro, B. (2023) · ACM SIGCOMM Conference (SIGCOMM) · doi:10.1145/3603269.3604828 · Cited by 9
“where there has been much interest in using data to drive design optimizations [10, 22, 24, 28, 40, 47].”
BackgroundMethodQoE-driven HAS Live Video Channel Placement in the Media Cloud
Liu, J., Zhang, W., Huang, S., Du, H., & Zheng, Q. (2020) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2020.2999176 · Cited by 9
“Yet another paper [31] developed CS2P to leverage a data-driven approach to modele the state-oriented evolution of throughput.”
Wang, S., Lin, J., & Ye, F. (2024) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2024.3437455 · Cited by 8
“Afterwards, many research efforts have been devoted to improving the time-slotted throughput prediction for MPC, such as CS2P [1] and Oboe [23].”
BackgroundMethodEnhancing QoE of Adaptive Video Streaming by Generating Fine-Grained Throughput
An, C., Zhang, H., Kang, J., Liu, Z., Zhou, A., Liu, L., & Ma, H. (2024) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2024.3508262 · Cited by 8
“The most recent solutions have enhanced the rule-driven algorithms by machine learning or deep learning techniques [5], [8], [17], [24], [25], [26], [27], [28], [29], [30], i.e. , learning-driven ABR algorithms.”
ALSTM: An Attention-based LSTM Model for Multi-Scenario Bandwidth Prediction
Li, P., Jiang, X., Jin, G., Yu, Y., & Xie, Z. (2021) · IEEE International Conference on Parallel and Distributed Systems (ICPADS) · doi:10.1109/icpads53394.2021.00018 · Cited by 10
“Accurate bandwidth prediction helps to improve the transmission performance of the DASH [1-3].”
BackgroundAccurate Bandwidth Prediction for Real-Time Media Streaming with Offline Reinforcement Learning
Tan, Q., Lv, G., Fang, X., Zhang, J., Yang, Z., Jiang, Y., & Wu, Q. (2024) · ACM Multimedia Systems Conference (MMSys) · doi:10.1145/3625468.3652183 · Cited by 11
QoE-Aware Resource Allocation for Small Cells
Elgabli, A., Elghariani, A., Aggarwal, V., & Bell, M. R. (2018) · Globecom · doi:10.1109/glocom.2018.8647828 · Cited by 8
Mobile Internet Throughput Prediction using Machine Learning Techniques
Jha, H., & Vijayarajan, V. (2020) · International Conference on Smart Electronics and Communication (ICOSEC) · doi:10.1109/icosec49089.2020.9215436 · Cited by 9
“[9 ] A more heterogeneous approach to the problem is covered in the method of studying both the mobile data and Wi-Fi performance and using paused video frames and energy consumption as a point for prediction of the future throughput.”
MethodCross-layer Network Bandwidth Estimation for Low-latency Live ABR Streaming
Shende, C., Park, C., Sen, S. B., & Wang, B. (2023) · ACM Multimedia Systems Conference (MMSys) · doi:10.1145/3587819.3590990 · Cited by 8
“Non-LLL streaming based rate adaptation has been studied extensively [9, 25, 30, 32, 34, 35, 39, 44, 51, 52, 57, 59, 63, 65], leading to buffer-based, rate-based, learning-based, or hybrid schemes.”
BackgroundOASIS: Collaborative Neural-Enhanced Mobile Video Streaming
Jin, S., Zhu, R., Hassan, A., Zhu, X., Zhang, X., Mao, Z. M., Qian, F., & Zhang, Z.-L. (2024) · ACM Multimedia Systems Conference (MMSys) · doi:10.1145/3625468.3647610 · Cited by 11
Twist: A Multi-site Transmission Solution for On-demand Video Streaming
Wang, H., Zhang, R.-X., Li, C., Xue, Z., Peng, Y., Pang, X., Zhang, Y., Ren, S., & Shi, S. (2024) · Proceedings of the ACM on Networking · doi:10.1145/3656297 · Cited by 8
“For example, adaptive bitrate techniques [33, 44, 45, 48, 51] focus on optimizing the selection of bitrates and adapt to dynamic networks to improve user experience.”
BackgroundDemystifying Myths of MEC: Rethinking and Exploring Benefits of Multi-Access/Mobile Edge Computing
Iwai, T., & Nakao, A. (2018) · IEEE International Conference on Cloud Networking · doi:10.1109/cloudnet.2018.8549539 · Cited by 9
“The algorithms for inferring network environments in which devices are located and have been developed to determine optimal bit rate and peers based on them [24]–[26].”
BackgroundSlow but Steady: Cap-Based Client-Network Interaction for Improved Streaming Experience
Krishnamoorthi, V., Carlsson, N. F., & Halepovic, E. (2018) · International Workshop on Quality of Service · doi:10.1109/iwqos.2018.8624170 · Cited by 8
“The clientbased techniques attempt to address the issue using robust quality adaptation algorithms [3], [23], [11] and better bandwidth estimation [24].”
MethodCross-layer DASH-based multipath video streaming over LTE and 802.11ac networks
Deng, Z., Liu, Y., Liu, J., & Argyriou, A. (2021) · Multimedia Tools and Applications · doi:10.1007/s11042-020-10393-8 · Cited by 8
“Rate-based algorithms [18, 25] usually request video segments at the highest bitrate that networks are predicted to support.”
BackgroundTimely and Accurate Bitrate Switching in HTTP Adaptive Streaming With Date-Driven I-Frame Prediction
Feng, T., Qi, Q., Wang, J., Liao, J., & Liu, J. (2022) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2022.3165381 · Cited by 8
“[11] Y. Sun et al., “CS2P: Improving video bitrate selection and adaptation with data-driven throughput prediction,” in Proc.”
MethodReliable Video Streaming With Strict Playout Deadline in Multihop Wireless Networks
Al-Zubaidy, H., Fodor, V., Dán, G., & Flierl, M. (2017) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2017.2742399 · Cited by 10
“[13]–[15], transmission rate estimation [16]–[19], or both [20].”
BackgroundEvolving Mobile Cloud Gaming with 5G Standalone Network TelemetryPreprint
Wan, H., & Jamieson, K. (2024) · arXiv (Cornell University) · doi:10.48550/arxiv.2402.04454 · Cited by 8
“Big data ML network prediction algorithms have shown potential to improve video on demand [47, 58], but such studies aggregate broadband and wireless last hops, and so do not necessarily directly point to improvements for game streaming.”
BackgroundUser Preference-Based Dynamic Optimization of Quality of Experience for Adaptive Video Streaming
Feng, Z., Liu, Y., & Zhang, H. (2025) · Electronics · doi:10.3390/electronics14153103 · Cited by 7
Battery-aware rate adaptation for extending video streaming playback time
Ahmad, H., Saxena, N., Roy, A., & De, P. (2018) · Multimedia Tools and Applications · doi:10.1007/s11042-017-5603-z · Cited by 7
“Both [46] and [41] perform adaptation based on throughput prediction.”
BackgroundAVIRA: Enhanced Multipath for Content-aware Adaptive Virtual Reality
da SILVA, F. R., Togou, M. A., & Muntean, G. (2020) · International Wireless Communications and Mobile Computing Conference (IWCMC) · doi:10.1109/iwcmc48107.2020.9148293 · Cited by 7
“Although [22], [23], [24] sample or measure " low level " network features, they use off-line analysis and training mechanisms to implement their proposed solutions.”
BackgroundEmpowerment of Atypical Viewers via Low-Effort Personalized Modeling of Video Streaming Quality
Peroni, L., Gorinsky, S., Tashtarian, F., & Timmerer, C. (2023) · Proceedings of the ACM on Networking · doi:10.1145/3629139 · Cited by 10
Zhang, D., Wei, L., Shen, K., Zhu, H., Wang, D., & Wang, F. (2024) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2024.3396810 · Cited by 7
User Mapping Strategy in Multi-CDN Streaming: A Data-Driven Approach
Zhu, G., & Gu, W. (2021) · IEEE Internet of Things Journal · doi:10.1109/jiot.2021.3112561 · Cited by 7
“Faced with the challenge of serving users distributed across widespread geographic locations, content providers deploy multi-CDNs in different regions to improve their service quality [29], [30].”
BackgroundImproving robustness of learning-based adaptive video streaming in wildly fluctuating networks
Lin, J., & Wang, S. (2023) · IEEE International Conference on Multimedia and Expo (ICME) · doi:10.1109/icme55011.2023.00307 · Cited by 7
SJA: Server-driven Joint Adaptation of Loss and Bitrate for Multi-Party Realtime Video Streaming
Shen, K., Zhang, D., Zhu, Z., Zhang, L., Wang, F., & Wang, D. (2023) · IEEE Conference on Computer Communications (INFOCOM) · doi:10.1109/infocom53939.2023.10229083 · Cited by 7
“[38] leveraged a data-driven approach to construct a Hidden-Markov-Modelbased midstream predictor to model the stateful evolution of throughput.”
QuDASH: Quantum-Inspired Rate Adaptation Approach for DASH Video Streaming
Wei, B., Song, H., Nakamura, M. M., Kimura, K., Togawa, N., & Katto, J. (2023) · IEEE Access · doi:10.1109/access.2023.3326326 · Cited by 7
“The ABR methods in existing literatures can be mainly classified into three kinds: (1) Rate-based [10], [11], (2) Buffer-based [12], [13] and (3) Hybrid [14], [15], [16].”
MethodZhao, L., Cui, Y., Jia, Y., Zhang, Y., & Nahrstedt, K. (2025) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2024.3521820 · Cited by 7
Learning on the Fly: An RNN-Based Online Throughput Prediction Framework for UAV Communications
Jiang, Y., Nihei, K., Li, J., Yoshida, H., & Kanetomo, D. (2020) · IEEE International Conference on Communications workshops · doi:10.1109/iccworkshops49005.2020.9145220 · Cited by 7
“For example, Yoshida et al. [7] model the throughput as a combination of stationary and non-stationary processes based on historical data, while Sun et al. [37] build a hidden Markov model for prediction, which should be executed in an offline manner.”
MethodCPU Microarchitectural Performance Characterization of Cloud Video Transcoding
Chen, Y., Zhu, J., Khan, T. A., & Kasikci, B. (2020) · IEEE International Symposium on Workload Characterization · doi:10.1109/iiswc50251.2020.00016 · Cited by 7
“Adaptive video streaming services tune video transcoding parameters to generate videos of different quality [18], [28], [38].”
Jin, H., Wang, Q., Li, S., & Chen, J. (2020) · International Conference on Control and Automation · doi:10.1109/icca51439.2020.9264312 · Cited by 7
“In 2016 and 2017, some machine learning-based ABR algorithms have emerged, a throughput prediction system CS2P proposed in [7] can be considered as the work of indirect ABR decision making by using machine learning algorithms.”
MethodMachine Learning Empowered Content Delivery: Status, Challenges, and Opportunities
Lu, Z., Gai, K., Duan, Q., Choo, K. R., Li, J., Wu, J., & Xu, Y. (2020) · IEEE Network · doi:10.1109/mnet.011.2000141 · Cited by 7
“The authors [12] also demonstrate that DNN appears to be better than the other techniques when predicting data-driven approach is designed to build a practical performance evaluation model for CDN cache server groups (the server nodes of CDN).”
PSTile: Perception-Sensitivity-Based 360$^\circ$ Tiled Video Streaming for Industrial Surveillance
Tu, J., Chen, C., Yang, Z., Li, M., Xu, Q., & Guan, X. (2023) · IEEE Transactions on Industrial Informatics · doi:10.1109/tii.2022.3216812 · Cited by 7
“CS2P [26] uses a hidden Markov model for throughput prediction to improve video bitrate adaptation.”
MethodNaresh, M., Saxena, P., & Gupta, M. (2023) · IEEE International Symposium on a World of Wireless, Mobile and Multimedia Networks (WoWMoM) · doi:10.1109/wowmom57956.2023.00024 · Cited by 7
“The rate-based algorithms [17] predict the future chunk’s bitrate as the maximum supported bitrate based on available network bandwidth and chunk history and the buffer-based algorithms predict based on the client’s buffer occupancy [18], [20].”
BackgroundMethodStreaming Video Optimization in Mobile Communications
Han, S., Ma, H., Chen, D., Wang, Y., Wu, Y., & Zhang, P. (2018) · IEEE/CIC International Conference on Communications in China (ICCC) · doi:10.1109/iccchina.2018.8641093 · Cited by 7
Network-Assisted Neural Adaptive Naked-Eye 3D Video Streaming Over Wireless Networks
Liu, Y., He, W., Wang, Y., & Yang, H. (2019) · IEEE Access · doi:10.1109/access.2019.2944437 · Cited by 7
“Most of the existing ABR algorithms are based on a fixed rule, or consider the estimated network throughput [33], [34], or consider the player buffer size [35], or comprehensively take these two quantities [36] into account.”
BackgroundFastconv: Fast Learning Based Adaptive BitRate Algorithm for Video Streaming
Meng, L., Zhang, F., Bo, L., Lu, H., Qin, J., & Han, J. (2019) · Globecom · doi:10.1109/globecom38437.2019.9013152 · Cited by 7
“The state-of-the-art ABR algorithms mostly make decisions based on network throughput ("rate-based" algorithms) [8], [9], buffer occupancy ("buffer-based" algorithms) [10], [11], or the combination of them both [12], [13].”
BackgroundEnergy efficient fuzzy-based DASH adaptation algorithm
Mowafi, M. Y., Taqieddin, E., & Al-Dahoud, H. (2020) · Digital Communications and Networks · doi:10.1016/j.dcan.2020.04.006 · Cited by 7
Optimizing Quality of Experience for Long-Range UAS Video Streaming
Shirey, R., Rao, S., & Sundaram, S. (2021) · International Workshop on Quality of Service · doi:10.1109/iwqos52092.2021.9521330 · Cited by 7
“HMM: Recent work [36] has shown that network throughput follows different states (with each state corresponding to different throughput distributions).”
BackgroundData-Driven Evaluation of Anticipatory Networking in LTE Networks
Bui, N., & Widmer, J. (2017) · International Teletraffic Congress · doi:10.23919/itc.2017.8064338 · Cited by 6
“[19], [20] propose a throughput prediction solution based on clustering and hidden Markov models.”
BackgroundCOREL: Constrained Reinforcement Learning for Video Streaming ABR Algorithm Design Over mmWave 5G
Hu, X., Ghosh, A., Liu, X., Zhang, Z.-L., & Shroff, N. B. (2023) · IEEE International Workshop Technical Committee on Communications Quality and Reliability (CQR) · doi:10.1109/cqr59928.2023.10317803 · Cited by 6
“On the other hand, throughput-based algorithms [10], [11] focus on matching the video bitrate to the estimated network throughput.”
BackgroundOptimizing Video Conferencing QoS: A DRL-based Bitrate Allocation Framework
Ko, K., Ryu, S., Tu, N. V., & Hong, J. W. (2024) · IEEE/IFIP Network Operations and Management Symposium (NOMS) · doi:10.1109/noms59830.2024.10575889 · Cited by 6
JSFBA: Joint Segment and Frame Bitrate Adaptation for Real-Time Video Analytics
Wang, Y., Cheng, S., Gao, N., Bi, T., & Jiang, T. (2025) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2025.3526867 · Cited by 6
“Sun et al. [7] developed CS2P, which employs a data-driven approach to learn clusters of similar video sessions and builds a hidden Markov model to reason about the state evolution of bandwidth.”
Learning Robust Adaptive Bitrate Algorithms with Adversarial Inverse Reinforcement Learning
Yi, L., & Qin, Y. (2025) · Chinese Journal of Electronics · doi:10.23919/cje.2024.00.202 · Cited by 6
Layer-Assisted Adaptive Video Streaming
Nasrabadi, A. T., & Prakash, R. (2018) · doi:10.1145/3210445.3210454 · Cited by 6
“Sun et al. [18] have shown that similar video streaming sessions share similar throughput patterns.”
BackgroundHybrid Control-Based ABR: Towards Low-Delay Live Streaming
Wang, B., Ren, F., & Zhou, C. (2019) · IEEE International Conference on Multimedia and Expo (ICME) · doi:10.1109/icme.2019.00135 · Cited by 6
Throughput Prediction-Enhanced RL for Low-Delay Video Application
Liu, Y., Zhang, C., Du, J., & Qiu, T. (2022) · International Conference on Mobility, Sensing and Networking (MSN) · doi:10.1109/msn57253.2022.00119 · Cited by 6
“The hybrid ABR algorithm [6], [7] considers throughput and buffers together.”
MethodCANE: A Cascade Control Approach for Network-Assisted Video QoE Management
Hosseinzadeh, M., Shankar, K., Apostolaki, M., Ramachandran, J., Adams, S. E., Sekar, V., & Sinopoli, B. (2023) · IEEE Transactions on Control Systems Technology · doi:10.1109/tcst.2023.3267716 · Cited by 8
“, tens of seconds) [30], [56], it is plausible to assume that the available bandwidth can be predicted for a short future horizon.”
BackgroundSelf-citationDemystifying Reinforcement Learning in Time-Varying SystemsPreprint
Hamadanian, P., Schwarzkopf, M., Sen, S., & Alizadeh, M. (2022) · arXiv (Cornell University) · doi:10.48550/arxiv.2201.05560 · Cited by 7
“The Markov chain generates a coarse bandwidth trace, which prior work shows is an appropriate model for TCP through-put (Sun et al., 2016).”
BackgroundTowards Impact of Chunk-Level Characteristics on Mobile Live Streaming Performance
Zhang, T., Tang, Z., Bao, J., & Ren, F. (2022) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2022.3207591 · Cited by 5
“According to [39], [40], the evolution of throughput within a session exhibits piecewise stationary characteristics.”
BackgroundLearning-Based Transport Control Adapted to Non-Stationarity for Real-Time Communication
Du, J., Zhang, C., Tang, T., & Qu, W. (2024) · International Workshop on Quality of Service · doi:10.1109/iwqos61813.2024.10682855 · Cited by 5
Fine-grained LTE radio link estimation for mobile phones
Bui, N., Michelinakis, F., & Widmer, J. (2018) · Pervasive and Mobile Computing · doi:10.1016/j.pmcj.2018.07.001 · Cited by 5
Towards secure and network state aware bitrate adaptation at IoT edge
Zeng, Z., Che, H., Miao, W., Huang, J. L., Tang, H., Zhang, M., & Zhang, S. (2020) · Journal of Cloud Computing Advances Systems and Applications · doi:10.1186/s13677-020-00189-4 · Cited by 5
“[35] systematically quantify throughput predictability using a large-scale dataset and propose a Hidden-Markov-Model based throughput predictor to enhance bitrate selection.”
ABRaider: Multiphase Reinforcement Learning for Environment-Adaptive Video Streaming
Choi, W., Chen, J., & Yoon, J. (2022) · IEEE Access · doi:10.1109/access.2022.3175209 · Cited by 5
“ML-based algorithms are classified into reinforcement learning-based [9], [15] and supervised learning-based approaches [11], [12].”
BackgroundMethodInfluentialGeo-Edge: Geographical Resource Allocation on Edge Caches for Video-on-Demand Streaming
Zhang, Y., Bian, K., Tuo, H., Cui, B., Song, L., & Li, X. (2018) · International Conference on Big Data Computing and Communications (BigCom) · doi:10.1109/bigcom.2018.00037 · Cited by 6
“We also collect bandwidth consumption traces of users throughout the days, and group the similar sessions [23].”
MethodOptimizing QoE of Multiple Users over DASH: A Meta-learning Approach
Huo, L., Wang, Z., Xu, M., Ding, Z., & Tao, X. (2019) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp.2019.8683246 · Cited by 5
“To solve the above challenges, several DASH adaptation approaches have been investigated such as [3–7].”
MethodECON: Modeling the network to improve application performance
Cao, Y., Nejati, J., Balasubramanian, A., & Gandhi, A. (2019) · ACM Internet Measurement Conference (IMC) · doi:10.1145/3355369.3355578 · Cited by 5
Q2ABR: QoE‐aware adaptive video bit rate solution
Amour, L., Souihi, S., Mellouk, A., & Mushtaq, M. S. (2019) · International Journal of Communication Systems · doi:10.1002/dac.4204 · Cited by 5
Modeling and Analyzing Live Streaming Performance
Zhang, T., Ren, F., & Wang, B. (2020) · International Workshop on Quality of Service · doi:10.1109/iwqos49365.2020.9212876 · Cited by 5
“According to [20], the evolution of throughput within a session exhibits piecewise stationary characteristics.”
BackgroundThroughput Prediction on 60 GHz Mobile Devices for High-Bandwidth, Latency-Sensitive Applications
Aggarwal, S., Kong, Z., Ghoshal, M., Hu, Y., & Koutsonikolas, D. (2021) · Lecture notes in computer science · doi:10.1007/978-3-030-72582-2_30 · Cited by 5
“, [15,41,47], became the state-of-the-art approach to ABR, as they combine the use of both throughput prediction and playback buffer occupancy.”
BackgroundMUABR: Multi-user Adaptive Bitrate Algorithm based Multi-agent Deep Reinforcement Learning
Yuan, H., Lu, H., Meng, L., & Liu, M. (2022) · International Conference on Communications · doi:10.1109/icc45855.2022.9839087 · Cited by 5
“Festive [3] and CS2P [4] are both rate-based.”
BackgroundDeep-Reinforcement-Learning-based User-Preference-Aware Rate Adaptation for Video Streaming
Lu, L., Xiao, J., Ni, W., Du, H., & Zhang, D. (2022) · IEEE International Symposium on a World of Wireless, Mobile and Multimedia Networks (WoWMoM) · doi:10.1109/wowmom54355.2022.00061 · Cited by 6
PAR: Improving Video Bitrate Adaptation via Payload-Aware Throughput Prediction
Pei, J., An, C., Zhou, A., Liu, L., & Ma, H. (2022) · IEEE International Conference on Multimedia and Expo (ICME) · doi:10.1109/icme52920.2022.9860000 · Cited by 5
“In general, the present ABR approaches can be classified into three categories: (i) rate-based [8, 9], which predicts available network capacity and selects the maximum bitrate that is not exceeding the estimated throughput; (ii) buffer-based [6, 7], which considers the current buffer occupancy and…”
BackgroundOnline Safety Assurance for Deep Reinforcement LearningPreprint
Rotman, N. H., Schapira, M., & Tamar, A. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2010.03625 · Cited by 5
“Example application domains include video streaming [2, 27, 49], traffic management [12, 57], resource scheduling [28], packet clas-sification [26], caching [22], and congestion control [62, 20, 32].”
BackgroundVariable-Sized Tiling and Rate Allocation for 360° Video Streaming
Su, Q., Huang, J., Li, W., Zhong, P., Gao, F., Li, Z., Liu, J., Zhou, S., Lu, B., Song, Z., et al. (2025) · IEEE Transactions on Vehicular Technology · doi:10.1109/tvt.2024.3524836 · Cited by 4
“CS2P[26]usesa data-drivenapproachtoanalyzethesimilarsession,andemploys thehiddenMarkovdatamodeltopredictfuturethroughputto choosethesuitablebitrate.”
Kan, N., Li, C., Jiang, Y., Dai, W., Zou, J., Xiong, H., & Toni, L. (2025) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2025.3596636 · Cited by 11
“…rate adaptation performance with a more accurate throughput prediction carried out by machine learning methods, such as the hidden-Markov models [17], probabilistic NNs with supervised learning [10], [18], [19], Bayesian NN-based prediction [11] and deep reinforcement learning (DRL) techniques…”
Agua: A Concept-Based Explainer for Learning-Enabled Systems
Patel, S., Han, D., Narodytska, N., & Jyothi, S. A. (2025) · ACM SIGCOMM Conference (SIGCOMM) · doi:10.1145/3718958.3754341 · Cited by 5
Energy and QoE Optimization for Mobile Video Streaming with Adaptive Brightness Scaling
Liu, D., Qian, C., Rong, H., Zhou, S., Xiang, C., & Jiang, H. (2024) · ACM Transactions on Sensor Networks · doi:10.1145/3670999 · Cited by 4
“However, existing ABR modules select video bitrate only based on the estimated network throughput [37] or buffer occupancy [19, 33], without regard to the actual context environment.”
PIRA: Pan-CDN Intra-video Resource Adaptation for Short Video Streaming
Qiao, C., Liu, T., Zhang, Y., Fan, Z., Xie, P., Wang, Z., & Liu, L. (2025) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/3746027.3754764 · Cited by 4
Self-citationBOLA360 : Near-optimal View and Bitrate Adaptation for 360-degree Video Streaming
Zeynali, A., Sahebdel, M., Hajiesmaili, M., & Sitaraman, R. K. (2025) · ACM Transactions on Multimedia Computing Communications and Applications · doi:10.1145/3785137 · Cited by 4
“This long period leads to noisy prediction of head position probabilities at the time of download, resulting in reduced QoE for the user.”
Session Throughput Prediction for Internet Videos
Li, Z., Kâafar, M. A., & Xie, A. G. (2016) · IEEE Communications Magazine · doi:10.1109/mcom.2016.1600386cm · Cited by 4
Machine Learning Based Transportation Modes Recognition Using Mobile Communication Quality
Kawakami, W., Kanai, K., Wei, B., & Katto, J. (2018) · IEEE International Conference on Multimedia and Expo (ICME) · doi:10.1109/icme.2018.8486560 · Cited by 4
“The authors of papers [3] proposed the prediction method that adopts Hidden Markov Model (HMM).”
MethodVideo data wireless transmission method based on cross-layer bitrate adaptation and error control
Yuan, J. (2019) · Multimedia Tools and Applications · doi:10.1007/s11042-019-7417-7 · Cited by 4
Machine Learning and Data Networks: Perspectives, Feasibility, and Opportunities
Lozada-Yánez, R., Molina-Granja, F., Yánez, P. E. L., & Yungán, J. I. G. (2020) · Advances in intelligent systems and computing · doi:10.1007/978-3-030-45691-7_26 · Cited by 5
Closed-Loop Clustering-Based Global Bandwidth Prediction in Real-Time Video Streaming
Afshar, S., Razavi, R., & Moshirpour, M. (2025) · IEEE Transactions on Machine Learning in Communications and Networking · doi:10.1109/tmlcn.2025.3551689 · Cited by 7
Energy-Efficient Predictive HTTP Adaptive Streaming in Mobile Cellular Networks
Tao, L., Gong, Y., Jin, S., & Zhao, J. (2018) · IEEE Transactions on Vehicular Technology · doi:10.1109/tvt.2018.2868791 · Cited by 4
“Recent studies [10]–[19], [21], [23]–[26], [31], [32] show that mobile users experience pronounced slow-fading, and the slow-fading trends can be predicted by mobility-pattern and localization.”
BackgroundExperience-Centric Mobile Video Scheduling Through Machine Learning
Chandrasekhar, V., Heng, Y., Cho, J., Lee, J., Zhang, J., & Andrews, J. G. (2019) · IEEE Access · doi:10.1109/access.2019.2933273 · Cited by 4
“The former involves Adaptive Bit Rate (ABR) algorithms that either learn [13] or apply control rules [14]–[16] for making bit rate decisions.”
BackgroundImproving Flow Scheduling Scheme With Mix-Traffic in Multi-Tenant Data Centers
Wang, S., Zhang, J., Huang, T., Pan, T., Liu, J., & Liu, Y. (2020) · IEEE Access · doi:10.1109/access.2020.2984277 · Cited by 6
“This idea is also used by CS2P [26] that designed for predict throughput of flows.”
MethodUnderstanding and Improving User Engagement in Adaptive Video Streaming
Qiao, C., Wang, J., Wang, Y., Liu, Y., & Tuo, H. (2021) · International Workshop on Quality of Service · doi:10.1109/iwqos52092.2021.9521277 · Cited by 4
Yanev, M., McQuistin, S., & Perkins, C. S. (2022) · International Workshop on Network and Operating Systems Support for Digital Audio and Video (NOSSDAV) · doi:10.1145/3534088.3534347 · Cited by 4
RetractedAdaptive Video Streaming Based on Learning Intrinsic Reward
Feng, Y., Wang, Y., Liu, H., Cong, L., & Liu, Y. (2022) · IEEE International Symposium on Broadband Multimedia Systems and Broadcasting (BMSB) · doi:10.1109/bmsb55706.2022.9828616 · Cited by 4
Samba, A., Busnel, Y., Blanc, A., Dooze, P., & Simon, G. (2018) · Computer Networks · doi:10.1016/j.comnet.2018.09.002 · Cited by 4
“[23] divide the session into periods of 6 s (epochs) and use a Hidden Markov Model (HMM) to predict the throughput of each period.”
MethodThe Cask Effect of Multi-source Content Delivery: Measurement and Mitigation
Chen, X., Zhao, M., Yang, X., Li, Z., Liu, Y., Li, Z., & Liu, Y. (2019) · IEEE International Conference on Distributed Computing Systems (ICDCS) · doi:10.1109/icdcs.2019.00034 · Cited by 4
“CS2P achieves improvement on overall QoE by using a data-driven model to learn clusters of similar sessions, an initial throughout predictor and a Hidden-Markov-Model [6].”
BackgroundMethodBandwidth Constrained Holographic Telepresence with 3D Model Reconstruction
Chen, Y.-X., Wang, C.-H., Yang, D.-N., & Liao, W. (2019) · Globecom · doi:10.1109/globecom38437.2019.9013866 · Cited by 4
“Sun et al. [10] leveraged the data-driven QoE prediction to design a bitrate selection and adaptation scheme.”
MethodPaCs: Playing Time-Aware Chunk Selection in Short Video Preloading
Fu, S., Yang, G., Yao, Z., Tan, S., Shan, Y., & Jiang, W. (2024) · Electronics · doi:10.3390/electronics13244864 · Cited by 3
DiVE: Differential Video Encoding for Online Edge-assisted Video Analytics on Mobile Agents
Shen, J., Zhu, H., Zhang, L., Li, Y., Chang, S., Wu, J., & Guo, M. (2025) · IEEE International Conference on Distributed Computing Systems (ICDCS) · doi:10.1109/icdcs63083.2025.00032 · Cited by 3
NeuroBA: Neuro-Symbolic Bitrate Adaptation for IRS-Aided Mobile Video Streaming
Chen, M., Xiao, W., Liu, A., Farouk, A., Chen, M., Niyato, D. T., Song, H., & Leung, V. C. M. (2026) · IEEE Transactions on Networking · doi:10.1109/ton.2025.3650266 · Cited by 3
“CS2P [23] clusters clients based on similarity features and models predictions as discrete Markov processes.”
Fine-grained LTE radio link estimation for mobile phones
Bui, N., Michelinakis, F., & Widmer, J. (2017) · IEEE International Symposium on a World of Wireless, Mobile and Multimedia Networks (WoWMoM) · doi:10.1109/wowmom.2017.7974286 · Cited by 3
“For instance, exploiting achievable rate prediction to optimize mobile applications [4]–[6] requires some information exchange between mobiles and base stations so that current decisions (e.”
BackgroundDemystifying the QoS and QoE of Edge-hosted Video Streaming Applications in the Wild with SNESet
Li, Y., Deng, G., Bai, C., Yang, J., Wang, G., Zhang, H. H., Bai, J., Yuan, H., Xu, M., & Wang, S. (2023) · Proceedings of the ACM on Management of Data · doi:10.1145/3626723 · Cited by 5
“For instance, adaptive bitrate selection (ABR) [96, 114, 117] requires studying the impact of different factors on network throughput or transmission time to determine the optimal chunk bitrate.”
Li, Z., Xu, C., Xiao, H., Fang, C., Zhong, L., Yang, S., & Muntean, G. (2024) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2024.3510036 · Cited by 3
“Throughput-Based Mechanism Algorithms : An example is the Throughput [35] algorithm, which relies on measured network throughput to adjust the video quality dynamically.”
Video Streaming with Kairos: An MPC-Based ABR with Streaming-Aware Throughput Prediction
Zhong, Z., Liu, M., Yang, L., Wang, Y., Xu, Y., & Hwang, J. (2025) · doi:10.1145/3712678.3721886 · Cited by 3
Liu, S., Wang, S., Ye, F., & Wu, Q. (2025) · IEEE Transactions on Green Communications and Networking · doi:10.1109/tgcn.2025.3598015 · Cited by 7
“For example, RobustMPC [9] leverages the harmonic mean of past throughput samples for predictions, while algorithms like CS2P [21] and Oboe [22] enhance throughput prediction for better decision-making.”
Metadata
- Authors
- Sun, Y., Yin, X., Jiang, J., Sekar, V., Lin, F., Wang, N., Liu, T., & Sinopoli, B.
- Affiliations
- ICT/CAS
- Capital Medical University (CN)
- iQIYI (China) (CN)
- Venue
- Proceedings of the 2016 ACM SIGCOMM Conference
- Publisher
- ACM
- Published
- 2016-08-22
- Pages
- 272-285
- OpenAlex
- W2482797934
- Cited by
- 407
- References
- 38
- Citations by year
- 2026: 20 · 2025: 25 · 2024: 40 · 2023: 43 · 2022: 40 · 2021: 57 · 2020: 55 · 2019: 50 · 2018: 54 · 2017: 21 · 2016: 2
- FWCI
- 36.18
- Type
- conference paper
- Language
- en
- Open access
- bronze
- doi.org
- Locations
- ACM SIGCOMM Conference (SIGCOMM) (publishedVersion)
- Funders
- National Science Foundation (1345305)
- National Natural Science Foundation of China (2012CB315802, 61379133, 61133015)
- National Key Research and Development Program of China (2012CB315802)
- Topics
- Physical Sciences › Computer Science › Computer Networks and Communications › Network Traffic and Congestion Control
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Image and Video Quality Assessment
- Physical Sciences › Computer Science › Computer Networks and Communications › Caching and Content Delivery
- Keywords
- throughput prediction, bitrate adaptation, video streaming, quality of experience, hidden Markov models, session clustering
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.