What are Recursive Congestion Signals in RINA ? Well this was a topic presented by the University of Oslo and PRISTINE researcher David Hayes, (co-authored by Peyman Teymoori, and Michael Welzl) at the recent IETF95 meeting in Buenos Aires.
In the IRTF research group, Internet Congestion Control (ICCRG) David took the time to present work on congestion control in recursive network architectures, highlighting the stability of chained and stacked congestion controllers, an overview of congestion signals and how they influence performance and a possible stable scalable new type(s) of congestion control.
There were a couple of questions noted in the meeting minutes such as
- ….. I did not see layered congestion protocols in the Internet. Is this real?
- Davids response: There are proxies.
- ….. Chaining is different to just chaining TCP – the signals from deep in the network needs to control the original source.
- Davids response: If it is a controlled source, rather than admission control.
- …… There could still be interactions if they operate on similar timescales. Coupling between loops is important.
- Davids response: Yes, one thing I didn’t mention though was that lower layers perform congestion control on aggregates.
- …… The interaction with chains is an important area.
Outside of the session minutes it was also good to hear that some of the work presented here is not just necessary for RINA but has wider applicability the key thing from the whole session is that no one said this was MAD, and so some comfort was taken from this.
The slides from the session are viewable on SlideShare or can be downloaded from the IETF 95 proceeding area.