A control-theoretic simplification of adaptive bitrate (ABR) video streaming

TL;DR

This paper introduces a control-theoretic approach to simplify adaptive bitrate streaming by using flatness-based control, closed-loop mitigation of network disturbances, and an algebraic bandwidth estimation method, improving QoE.

Contribution

It presents a novel, simple control strategy for ABR streaming, combining flatness-based control with model-free techniques and a new bandwidth estimation method.

Findings

Effective buffer and bitrate control via flatness-based strategy.

Closed-loop control mitigates network fluctuations.

Algebraic bandwidth estimation handles capacity variations.

Abstract

Adaptive bitrate streaming (ABR) over the HyperText Transfer Protocol (HTTP), which raises numerous delicate questions, is nowadays almost the only approach to video streaming. This paper presents elementary solutions to three key issues: 1) A straightforward feedforward control strategy for the bitrate and the buffer level via flatness-based control. 2) Closing the loop permits mitigating unavoidable mismatches and disturbances, such as Internet fluctuations. This is adapted from the new HEOL setting, which mixes model-free and flatness-based controls. 3) An easily implementable closed-form estimate of the bandwidth via algebraic identification techniques is derived, perhaps for the first time. It permits handling severe variations in channel capacity. Several computer experiments and metrics for evaluating the Quality of Experience (QoE) are displayed and discussed.