there was a seminar organized earlier this month in Paris. The goal of that seminar was to gather various research teams working (or considering to work) around RIOT at Inria.
Securing IoT is important now. (see slides from C. Bormann)
Formally verified open source IoT software is desirable and some is already available today (see slides from K. Bhargavan)
Public key crypto is the hardest part of IoT crypto, but significant performance improvements are on the horizon (see slides from B. Smith)
Industrial IoT deployments are already happening. (see slides from T. Watteyne)
Object security, and content-centric networking can be used for security. (see slides from M. Enguehard)
Need public experimental infrastructure & open data for IoT. (see slides from E. Fleury)
Large open source project need tools/languages to (i) flexibly express how code should be and (ii) automate code transformation (see slides from J. Lawall)
OS adaptation and programming abstractions are needed for intermittently powered IoT devices, and peripherals restoration is the hard part (see slides from G. Salagnac)
IoT software components and component dynamic updates are desirable. (see slides from J. Bourcier)
If you have comments or questions, just ask here and/or contact the authors directly!
Thanks for the links! I was very interested to see Thomas’s slides on OpenWSN “fresh” [1]. It would be valuable to have a 6TiSCH implementation in RIOT. Are there any more details available on this project?
Hi Ken,
there used to be OpenWSN support in RIOT but due to a missing maintainer on our side (we did not find the time to adapt it to our rapidly changing APIs for timers and networking) we had to remove it for now [1].
Thanks for the reference, Martine. I should have been clearer. I was particularly referring to slide 18 on page 9. It looks to be a new initiative to support this integration.
Hi Ken,
this is planned for the summer. We discussed with Thomas, and the rough plan was to propose this topic for the IETF hackathon [1] in Prague this summer, whereby some RIOTers and some OpenWSN people would attend and hammer it out. Until then, OpenWSN Fresh is getting ready AFAIK.
I would like to let you know about our recent work [1] on access control for the IoT. You can find a pdf version of this work here [2] and a reference RIOT-based implementation here [3]
The basic idea is that a trusted third party helps a Thing and an *authorized* user to establish an ephemeral (session-specific) encryption key. It is lightweight (in uses only HMAC and AES) and it provides Thing authentication.