In a thread that I lost, Peter asked:
since the IoT WG released it's first BCOP document [1] some time has passed. Several technologies discussed in the document – i.e. DPP, USP – have become mandatory for large carriers and can be found the majority of their roadmaps.
That's really news to me. I guess maybe this is about USP more than it is about DPP. I'm unware that DPP has become mandatory. As for the USP (TR-369), I'd like to know more: it's quite a fungible spec, and what part exactly has become mandatory, and what transports? In our BCOP, we said we needed a standard phone/browser <-> home gateway API such that we could have innovation at the edge. (Right now, it's a market vertical with a new app for every home gateway) So I'd love to have some discussion, maybe there are even presentations, about where we are here.
Recent talks with carriers revealed that all intend to increase their efforts in managing IoT devices in their customer's homes or start such efforts. All of them refer to Thread [2] and Matter [3] as the technologies of choice – which makes sense given that at these standards increase compatibility and are backed up by large players like Google, Amazon & Apple. Since Thread is based upon IPv6 it offers new options and possibilities to integrate with existing network infrastructure an standards. Nevetheless it also requires re-evaluation of threats and attack vectors in such networks.
yes, but I think that the carriers are gonna find that they have some serious privacy implications if they get involved. I think that it is gonna bite Google and Apple and Amazon too.
I would like to initiate work on a follow-up BCOP document that centers around the changes Thread and Matter do introduce to networks, but that also takes into account our findings that got incorporated into the fist BCOP document this WG has released. Especially the challenges regarding onboarding and security in general remain the same and would be needed to be evaluated in regards to the new technologies.
One thing that might be useful is to have a Thread and Matter (they are different and also dependant) reference networks that we could play with in person at a future RIPE meeting. Probably too soon for Rotterdam, but maybe in the fall. (Wherever that will be)
An IPv6 based IoT networking protocol offers also new possibilities for integrating IoT devices into networks and can change the way we look at such networks. This should also be part of such a document.
I would like to ask the members of this WG about their opinion about these considerations. And I also want to ask for volunteers to work an such a new BCOP document and would be glad to initiate and manage the process.
In general, I think that it's too soon to say anything useful in a new document in 2023, but that we should be spending some time understanding what's going on. -- Michael Richardson <mcr+IETF@sandelman.ca> . o O ( IPv6 IøT consulting ) Sandelman Software Works Inc, Ottawa and Worldwide