Monday, October 4, 2010

60 Ghz and the Future

In one of my earlier blogs, I pointed out that about 40 years ago 2 meters and 70 cm were basically uncharted areas. Now they are populated. Undoubtedly the future of ham radio is in our huge - virtually unused microwave allocations. They have the necessary bandspace to support wideband modes.

Jim, KC4BQK was first to repost this good video about 60 Ghz and the future of LANs. Blogger Craig Mathias from the Fairpoint Group dose a good job explaining 60 Ghz and the difference between WiGig and WiMedia. This could be something in the future that could include Ham Radio.

Check out his blog

Sunday, October 3, 2010

Linux Speech Recognition

I've been looking for Automated Speech Recognition for Linux for a while. I keep seeing Dragon Naturally Speaking on the selves locally.

I have read some discussions where people have gotten Dragon to work under Wine. But I really need something that works standard in and out style, so I can script things to work with it.

I have all kinds of ideas for two-way radio integration and Asterisk projects. If only something existed.

I have played with CMU Sphinx in the past, but it's not as developed as I'd like.

If anyone has experience in the area of Speech Recognition with Linux, I'd like to hear from you. Ideally I am looking for something that is under $200, and works as well as Dragon. It need not be open source, just Linux compatible.

Specifically if anyone has experience with Voxeo's Prophecy, or LumenVox, I am eager to hear more.

I doubt this online petition for a version of Dragon Naturally Speaking for Linux, will convince Nuance Communications, but I suppose it's worth a try.

Saturday, October 2, 2010

P25 Development

It's been a while since I reported on anything P25 related.

Much has been happening with the op25-dev group.

The code is still on the private sedition repo until they are ready to make an initial release and the current re-engineering of the code has been held up by a need to complete a thesis.

Apparently the plan is to re-home the repository on CGRAN or Google Code before the year end and make a public release at the same time.

If you are not familiar with the work going on, simply put; it's a Huge interoperability stride for narrowband digital radio.

An interesting suspected fork of this work is a package called the Digital Speech Decoder.

This already decodes C4FM off a sound card with a discriminator connection, and it error corrects.

If it was possible to take this a pipe it back out a sound card (without passing it to mbelib) to a transmitter in theory you could build a P25 repeater using a computer and a couple sound cards with the appropriate discriminator and varactor connections.

Jonathan Naylor, G4KLX has done this sort of thing with D-Star. And I found it most impressive.

From there the raw IMBE bit stream and headers could be IP encapsulated, sent
over the internet for the purpose of P25 repeater linking.

I don't know how many P25 groups are out there, but I'd encourage them to get the
word out to anyone with the potential programming talent.

And possibly place a bounty to help encourage this type of development.