Technical Summary
The RMCAT working group is developing congestion control schemes for use
with RTP. When multiple such congestion controlled RTP flows traverse the
same network bottleneck, combining their controls can improve the total
on-the-wire behavior in terms of delay, loss and fairness. This document
describes such a method for flows that have the same sender, in a way
that is as flexible and simple as possible while minimizing the amount of
changes needed to existing RTP applications. It specifies how to apply
the method for the NADA congestion control algorithm, and provides
suggestions on how to apply it to other congestion control algorithms.
Working Group Summary
The draft has been under development in the working group for some years.
Much of the time was taken waiting for the candidate congestion control
algorithms to stabilise, mapping the algorithms to the mechanisms given
in this draft, and deciding which congestion control algorithms should be
supported. The coupled congestion control algorithm itself has proved
reasonably stable.
The draft discusses how to apply coupled congestion control to NADA and
Google Congestion Control. The mapping to NADA is in the main body of the
draft, since NADA is nearing working group last call and believed stable.
The mapping for Google Congestion Control is in an appendix, since Google
Congestion Control is not yet finalised. There is no mapping for SCReAM
at this time, but one could be added later if there was interest in doing
so (nothing in SCReAM should prevent this).
Overall, the working group process has been relatively smooth, although
not rapid. The main issue of contention was the choice of congestion
control algorithm to which the mechanism should be applied - based on
the maturity of the candidate congestion control algorithms, and the
relative importance the authors of the candidates placed on coupled
congestion control.
Document Quality
The algorithm has been implemented in simulations and emulated testbeds.
This is appropriate for an experimental protocol of this type, and meets
the usual community evaluation standards for transport protocol research.
The draft has been reviewed by some authors of each candidate congestion
control algorithm, with Xiaoqing Zhu and Stefan Holmer providing detailed
reviews and advice on integration with the congestion control proposals.
The draft is well written, and the mechanism is clearly specified.
There is no need for MIB Doctor, Media Type, or other expert review,
since the proposed mechanism relies only on common RTP features and
parameters that can be directly measured by the end-point using the
mechanism.
Personnel
The document shepherd is Colin Perkins.
The responsible AS is Mirja Kühlewind.