A standard communication protocol will definitely enable better competitive market and will ultimately help advancing the technology. However, a standard could introduce security issue. I'm sure all these organizations have put in a reference architecture with security in mind. Question is whether there can be a better collaboration between different standard bodies so that there are at least a standard for communication between clusters following different standards.
The trouble with having a single standard communications protocol is that it may be too complex for some applications and perhaps not capable enough for others. I have difficulty with the idea that my IoT thermometer reading outside temperature every few minutes will use the same protocol as the IoT devices monitoring the real-time performance of a jet engine. My expectation is that different application domains will choose their own sets of standards to use. I agree that security needs to be a key part of any reference architecture, though.
There might be a role for government in standards that affect public safety, but I don'tthink that governments need to (or should) be involved in most commercial applications of the IoT. The people working in the field have a much better handle on the needs and challenges than anyone in government. And I trust market forces to work out a standard eventually.
Absolutely. I don't think it is a good idea to run full-blown TCP/IP stack on IoT devices. There are at least 2 reasons. Firstly, the footprint may not fit well. By increasing memory size, the power consumption becomes a challenge. Secondly, the longer the packet, the higher the power consumption on transmit.
@RichQ precisely! There was an active participation by NIST folks in TIA-TR50 committee as NIST was also involved in Smart Grid Interoperability Standards. NIST in my opinion is atypical of other goverment bodies that meddle in regulations!