Wave Fronts

Software Defined Radio Use Case

Smarter elimination of SSB adjacent signal interference

I've been a licensed radio amateur for 11 years and I'd only had one interest as far as application and that was HF radio contesting. HF is short for "high frequency" and covers about 1-30mhz with 10 amateur bands of frequencies within this range. HF contests involve trying to make more contacts in a fixed amount of time where each contact is short to very short, the jargon equivalent to "I'm Pete and my favorite color is violet." It's actually like "F5XY 59 5" spoken very quickly or "F5XY 5NN 5" by morse code ("N" is contester short hand for "9" as dah dit is shorter than dah dah dah dah dit).

But after moving house and deciding against another tower and "serious antenna system" I realized I needed a new interest. And I need a reason to be in the "radio room" (the back of a detached storage building) as this is where my daughers electronic tinker toy set is in position to keep all the bits and pieces from going to never land. I've settled on exploring something called a "software defined radio" and the rest of this is about my first application goal.

I recently took part in the ARRL phone(voice) "Sweepstakes" in which you can make a lot of contacts if you can send and receive signals well and can stand to pick out other voices in a cacaphony of noise. Really, a major phone contest is just weird. I wish I had an mpeg to link to to give a sense of what it's like. Basically you're soliticing contacts, saying effectively "please let me tell you my favorite color" while somebody just below you in terms of frequency and somebody else just above are doing the same, hammering your ears with interference relative to your goal.

With single side band (what we use for "phone" contests) it's like an old style telephone party line: all voices can be heard at the same time, but shifted up or down in tone. This is very different from AM and FM in which one signal steps on or completely displaces another. So with side band you've got somebody soliciting contacts below you but their audio is shifted up or down by the difference in radio frequency of their signal relative to yours and also your listening frequency. The net effect is that it's kind of like trying to carry on short conversations while Donald Duck and Tweety Bird are yammering in your ears. And sometimes the folks responding to their solicitations are also audible. In some cases the soliciting signal is louder than their responders, other times its visa versa or about the same loudness. The loudness is proportional to signal strength, ignoring automatic volume control factors.

So I started wondering if it might be possible to apply a computer to a situation where even state of the art amateur band receivers can't get the job done. That is, if I can hear and recognize the lower pitch and higher pitched Donald Duck and Tweety bird interference, couldn't a computer do it too? Computers with DSP (digital signal processing) are extremely effective for locating and notching out a single tone, such as is frequently caused by a continuous carrier produced by another station. So one moment you've got a killer tone in your ear and the next, after you push the "heterodyne filter button" it vanishes.

So why can't this scheme be generalized to cause Donald Duck to vanish from my ears? Even if I could just get the level down a little it would be great.

So this is my first goal: to try to arrange for "alien" audio components within my passband to be attenuated while leaving "friendly" audio tuned to my frequency alone. I have no idea if this is an impossibility or if it has been done already.

Unfortunately, the GNU Radio Universal Software Radio Peripheral I bought to explore this area seems to be defective. I haven't had the time to discuss this with its maker. I'm hoping it's something simple, or maybe something that is at least repairable, like a cracked solder joint.

And, in the meantime, I've broken through self-consciousness with the fumble-bumble way I do morse code (heavily depending on computer assistance), and I find CW contests to be so much easier than phone, I might put off this SSB-research for some time. And in a CW contest you can eat and drink while participating, something that's very hard to do when you're making a lot of contacts per unit time for hours on end.

Posted at 09:00PM Jan 06, 2006 by microwaves in Radio  |  Comments[2]