06.19.08

1900.5 – Maybe policy languages won’t kill the radio star

Posted in cognitive standards, language at 2:22 pm by JamesNeel

One of the things I harp on a lot is the impracticality of DARPA’s xG policy ontology solution as it has an unbounded order of complexity (see Table 13.1 in B. Fette, Cognitive Radio Technology). So I’ve been mostly ignoring the language standardization efforts (e.g., 1900.5 til now and MLM in the SDRF) because I thought they were never really going to get deployed.  

At the SDR Forum meeting in Portland, 1900.5 held a couple sessions, which I wasn’t originally too keen to sit in on and only ended up attending because a public safety sig CR session got canceled. While the actual session devolved into a ”what is a language?” exercise in intellectual onanism as I expected, I ended up speaking some with John Strassner (the lead on 1900.5) afterwards.

In addition to recommending a nice salmon dish, he said his vision for 1900.5 is a policy language that can be compiled and implemented on a radio. If that works out (and John seems to have the background for this to happen), then policy reasoning should be feasible to do in real-time on practical radios and my entire opinion of the utility of policy languages will have to change.

This was apparently the goal all along for the group, so perhaps I should’ve known this, but I previously had a very low opinion of the potential utility of a policy language so I wasn’t following closely – which I’ll have to change.