BOLA: Near-Optimal Bitrate Adaptation for Online Videos (1601.06748v3)

Abstract: Modern video players employ complex algorithms to adapt the bitrate of the video that is shown to the user. Bitrate adaptation requires a tradeoff between reducing the probability that the video freezes (rebuffers) and enhancing the quality of the video. A bitrate that is too high leads to frequent rebuffering, while a bitrate that is too low leads to poor video quality. Video providers segment videos into short segments and encode each segment at multiple bitrates. The video player adaptively chooses the bitrate of each segment to download, possibly choosing different bitrates for successive segments. We formulate bitrate adaptation as a utility-maximization problem and devise an online control algorithm called BOLA that uses Lyapunov optimization to minimize rebuffering and maximize video quality. We prove that BOLA achieves a time-average utility that is within an additive term O(1/V) of the optimal value, for a control parameter V related to the video buffer size. Further, unlike prior work, BOLA does not require prediction of available network bandwidth. We empirically validate BOLA in a simulated network environment using a collection of network traces. We show that BOLA achieves near-optimal utility and in many cases significantly higher utility than current state-of-the-art algorithms. Our work has immediate impact on real-world video players and for the evolving DASH standard for video transmission. We also implemented an updated version of BOLA that is now part of the standard reference player dash.js and is used in production by several video providers such as Akamai, BBC, CBS, and Orange.

• The paper introduces BOLA, a buffer-based Lyapunov algorithm that frames bitrate adaptation as a utility maximization problem for smoother video streaming.
• It achieves near-optimal performance by delivering 84–95% of the offline optimal utility through extensive simulations without depending on bandwidth forecasts.
• The algorithm's integration into dash.js and adoption by major entities like BBC and Akamai underscores its practical impact on adaptive streaming.

BOLA: Near-Optimal Bitrate Adaptation for Online Videos

In the contemporary landscape of internet video consumption, the challenge of providing optimal user experience is primarily navigated through adaptive bitrate (ABR) streaming. This methodology is pivotal in accommodating the heterogeneous nature of network bandwidth and device capabilities. The paper "BOLA: Near-Optimal Bitrate Adaptation for Online Videos," authored by Kevin Spiteri, Rahul Urgaonkar, and Ramesh K. Sitaraman, introduces a novel algorithm termed BOLA (Buffer Occupancy based Lyapunov Algorithm), which aims to maximize video quality while minimizing interruptions due to rebuffering.

Problem Formulation and Methodology

BOLA's foundation is rooted in framing the bitrate adaptation task as a utility maximization problem. The utility considers both the quality of experience (QoE) from high-bitrate video and the penalties from playback interruptions. Unlike bandwidth prediction-based approaches, BOLA employs Lyapunov optimization, focusing on buffer occupancy without estimating network bandwidth. This shift presents a theoretical sophistication that simplifies algorithm design, sidestepping potentially erratic bandwidth forecasts.

The authors rigorously prove that BOLA achieves a time-average utility additive factor of $O(1/V)$ from the optimal value, contingent on a control parameter $V$. This theoretical underpinning imbues BOLA with robust performance characteristics, ensuring that its practical implementation adheres closely to its theoretical predicted outcomes.

Empirical Validation

Through an extensive series of simulations utilizing both static and mobile bandwidth traces, BOLA was evaluated against leading bitrate adaptation algorithms. It exhibited marked improvements in utility, often surpassing state-of-the-art methods like PANDA and MPC under varying conditions. Specifically, BOLA demonstrated utility within 84–95% of the offline optimal, highlighting its efficacy in real-time applications without necessitating future bandwidth predictions, which are typically unattainable in live scenarios.

Deployment and Impact

A significant achievement of this research is BOLA's integration into dash.js, the reference implementation for MPEG-DASH, now utilized by major entities such as BBC and Akamai. This adoption underscores BOLA's practical value and adaptability within existing ABR frameworks. The inclusion of BOLA in production settings and its empirical success has implications for ongoing MPEG-DASH standard developments, reinforcing the relevance of buffer-based adaptation strategies.

Implications and Future Directions

The introduction of BOLA offers several practical and theoretical implications. Practically, it simplifies ABR implementations by reducing reliance on bandwidth estimation. Theoretically, it validates buffer occupancy as a pivotal metric in ABR strategy and paves the way for further innovations in non-predictive adaptation algorithms.

Future research could explore synergistic approaches combining BOLA with AI models, as the computational efficiency and theoretical guarantees of BOLA could complement the adaptability and learning capabilities of AI-driven algorithms. Additionally, extensions of BOLA to accommodate newer video streaming formats and more diverse viewing scenarios would further solidify its position in adaptive streaming research. Overall, BOLA signifies a significant step towards more reliable, efficient, and user-oriented video streaming experiences.