JT65/JT9 not impervious to false positives…

January 20, 2014 | Amateur Radio | By: Mark VandeWettering

I haven’t had a lot of time to operate JT65/JT9, my preferred digital mode at the moment, but I often leave my receiver hooked up and monitoring the bands for reception reports. My wet-noodle of an antenna can usually pull in signals from about 40 different counties in the span of a week of monitoring. I usually leave it running on 40m during the night, and 20m during the day. Sometimes I’ll toss in 15m just to see what’s going on. Here’s a screengrab of the last 24 hours, for instance:

full

If you want, you can go to the PSKReporter website and enter my callsign K6HX (or any other) and get a live report. In this map, the blue dots are signals that I received on 40m, the yellow ones are 20m, and the brownish ones are for 15m. If you look at these regularly, and over time, you can begin to get some idea of how propagation works by frequency and by time of day.

But look at that map closely, in particularly in the area up above Scandanavia, you should see a single yellow mark well above the Artic circle. A little poking indicates that this was for the rather unusual callsign JW4BHO, which comes from Svalbard. I’d never seen a radio spot from there, which is always exciting. I looked up JW4BHO on the callsign database on QRZ.com, and I felt a bit sadder. His callsign wasn’t registered: not a good sign. I also did a db search on PSKReporter to see if anyone else had spotted him. No one else had. Rats.

One thing bothered me though: the callsign and the marked for the reception report matched. JT65/JT9 signals both encode the Maidenhead gridsquare to indicate the position of the transmitter. If this was just random noise, then how did the location and the callsign match? That seemed very unlikely by chance.

Luckily, I still had WSJT-X up, and could scroll back and look at the reception report:

crop

Here you can see that WSJT-X reported a signal from JW4BHO to T07NAA from gridsquare OQ65, at a SNR of just -17. -17 isn’t a particularly weak signal, which surprised me a little, but the reception report is obviously bogus. T07 doesn’t even refer to a valid country. But somewhat oddly, the gridsquare OQ65 isn’t even the gridsquare marked on the map. For some reason, the PSKReporter website has decided to substitute a generic square marking the center of the country instead of using the actual gridsquare marked in the reception report. That seems odd to me.

And this isn’t the first time it happened to me, I’ve seen it a couple of times over the last month, including a fictious spot from Somalia.

Has anyone else had this problem?

Comments

Comment from John Smith
Time 1/21/2014 at 3:29 pm

It’s possible that it’s just unlicensed radio amateur using fake/random ID.

ps. I love your blog. And I would like to thank you for giving professionall RSS feed when its possible to read WHOLE article over RSS instad just snipped (this is very very very useful for people without pernament internet connection – I can just connect to wifi spot somewhere for 1 minuete per week and update all articles).

Thank you. John Smith

Comment from John Smith
Time 1/21/2014 at 3:30 pm

oops. sorry for typos and errors in previous message. please keep the rss in current form – its awsome. (you are the 1% of the internet doing RSS right)

Comment from Mark VandeWettering
Time 1/22/2014 at 8:24 am

Interesting theory but the fact that nobody else heard them strongly suggests that these are phantoms.

Comment from Eric T- N6SPP/KL
Time 2/1/2014 at 1:40 am

Even the JR1AQN @ 2325z might be false.. AT A dt of 1.8secs he is a border line decode..Maybe he is portable in CT. hi

I’ve used WSJT for meteor and tropo (+later versions for eme +hf)since 2001.. Versions 4.9,6,7 and 9.x contain a “call3 txt” folder with hundreds of call signs as you may know. This on board database tries to match partials(bits of call signs) and other criteria to form a valid decode out of the standard message formats.

Using “deep search”, “aggressive deep search” and adjusting the quasi squelch/decode threshold to ” -1″ can net a few false decodes.

It appears though, that WSJT-x contains a small temporary call3 txt folder.

So, I’m not quite sure how the false decodes happen in wsjt-x.. I have also gotten a few. I’ll leave that up to you guys- the programmers! hi. Maybe you can help me with an explanation.
(the MMMonVHF site updates the Call3txt list daily)

As John Smith mentioned, a stn can place any 4 or 6 digit gridsquare in the setup menu.

ps- i wk’d JW5E last sept on 20m phone @0052z.. He was at the NRRL Club stn and I was at our KL7AA club. Pointing the club’s tribander due North easily nets Eu and Scandinavia from KL7.

73, Eric n6spp-anchorage
(DeAnza Class of ’80 hi)