Decode or not decode - that is the question...

Having observed WSPR for a number of days (which of course qualifies me as an expert... ;-), I have noticed a puzzling thing. The decoding success does not correlate well with observed signal strength as observed on the trace (at least at this location).

That is, sometimes a very faint trace will decode well and often. But sometimes a strong signal decodes only occasionally (at this QTH ZL2FT is an example - but also other single-skip stations exhibit this as well).

Just when I thought it was a characteristic of strong local (<5000km), the same thing happened for further stations.

A case in point - I decoded GW4BVE on 17M (dist. ~17000km) on a trace which show a jump in frequency, but later on no decode on a signal just as strong and another even stronger (see blog entry "GW4BVE decode..." for link to screenshot).

I am guessing this is a characteristic of the strong forward error correction - where it basically operates under the principle "if in doubt - chuck it out..".

From my experience many years ago on HF narrowband there are times (epochs around tens of minutes) where a single tone is 'smeared' by effective path length variations and I have observed then that sometimes a strong signal will be 'smeared' while a weaker signal received on another path has a nice single trace. Maybe this is the effect here.

I did some experiments (about 8 years ago) with a technique I called Frequency Difference Keying (FDK) specifically to overcome this resulting in a decode in Minnesota, USA, of my 2mW signal here in VK into a mobile whip on the 22M ISM band.

Unfortunately I couldn't persuade anyone to write an application suitable for general use (a little beyond my capability at the time) and so I went on to invent the IFK (Incremental Frequency Keying) technique, which is now used in Jason and DominoEx. BTW - even though these programs were written by other parties, they kindly gave my credit for inventing the technique - a refreshing gesture on their part.

Now that I am retired (yippeeeeee!!!!) I might re-visit my FDK technique (I say my FDK technique - but I think NASA use it for deep space craft status..) to see if it might be useful. Need to dig out the code from many moons ago...