Validate ABR behavior in version 5.0.0 against a testbed

[dsilhavy]

Description

In version 5.0.0 of dash.js several changes are made to the throughput calculation and the ABR behavior in general. A first large PR has already been merged: #4237 .

The goal of this ticket is to evaluate the different options for throughput calculation together with the different ABR rules in dash.js against an ABR testbed. Focus should be set on the following high-level questions:

• Do we see any regression/worse behavior compared to dash.js 4.7.2?
• Which throughput calculation method performs the best/most accurate?
• Which combination of ABR rules performs the best for "standard" streaming and "low-latency" streaming?

The definition of "best" should be clarified here as well. For that reason, existing QoS metrics should be used such as bitrate, rebuffer rate, quality oscillations, latency etc.

[dsilhavy]

Current development status of v5 is reflected here: https://reference.dashif.org/dash.js/v5/samples/dash-if-reference-player/index.html

[dsilhavy]

@minhkstn to present first results at the face to face