|Re: semantic description in om2m [message #1490107 is a reply to message #1485447]
||Thu, 27 November 2014 23:51
| Mahdi Ben Alaya
Registered: November 2013
Hello Fesseha, |
OM2M is developed within the LAAS-CNRS laboratory by a team composed mainly of researchers. Internally, we have a version of OM2M including some advanced features related to our research topics where we are experimenting semantic models and reasoning, information centric networking, autonomic computing, and so on. At the same time, we are working on extending the open source version of OM2M with new plugins useful for developers such as interworking with zigbee, knx, mqtt and 6lowPan, and provide a nice front and back-end interfaces as well.
SSN is based on RDF, an XML based format so not really efficient for constraint devices. But at least, you can use SSN concepts and relations in another lightweight format. This will help you to reduce the quantity of new names to be used in your system, and so decrease the semantic gap. The main idea from using SSN here is the ability to do semantic reasoning on data by a more high level application. Actually we are using the Core Link Format from IETF to describe hosted resources withing constrained devices. I am not familiar with IPSO smart objects model, and I am not sure if it supports semantic. For me, IPSO is something similar to oBIX but I will try to investigate more this alternative. I agree with about XML overhead, that's why we are working on integrating JSON in OM2M.
OM2M is still in the incubation phase. All information about the roadmap and news will be detailed here: https://wiki.eclipse.org/OM2M#Project_Management. We are now updating some third libraries to complete the Eclipse validation process to go to our first release which will include json format, zigbee interworking proxy plugin and a nice charts interface.
Powered by FUDForum
. Page generated in 0.01575 seconds