Spotters using KiwiSDR

[Andy G7IZU]

Hi, it's my first time posting here so apologies if this has been asked before. 

I see a number of people using KiwiSDRs when reporting. Are these local deivces or are they using them remotely through the web, and how can you tell? If they're remote, is the location of the receiver being input in order to plot the end points properly on the map? 

Regards,

Andy.

[Sissy KM4QLB]

Great question, Andy!  When reception reports of my signal come from people halfway around the world, my elation often turns to frustration when I discover that the reporters are using remote access receivers like KiwiSDR.  I don't need PSK Reporter to know that the Internet has vast coverage; I'm trying to see what coverage my modest little QRP rig has.

PSK Reporter feature request:  Allow community members to flag suspected mismatces between reported grid square and antenna location.  Once enough people do that for a given reporter, their reports should be excluded from the map except when people click a box that essentially says 'show them to me anyway.'  PSK Reporter could let people remove themselves from that list by including something like "antenna is local" in their rig description.  Philip Gladstone is a smart guy and doesn't need my help in working out details, so I'll just leave the general suggestion here.

Regards,

  Sissy

[Luke Williams]

Andy,

If the report shows "KiwiSDR" as the rig, then it is probably being
reported by the KiwiSDR system directly and more than likely accurate.
KiwiSDR is all web-based and spotting is being done by the server, not
the browser. There's a project called DigiSkimmer that connects to
KiwiSDR and those could be run remotely from the KiwiSDR. That
situation might be more likely to be incorrectly configured, but that
is just speculation. There are also not that many of them reporting
compared to native KiwiSDR. They would be reported as "DigiSkimmer
0.35.1 KiwiSDR" or similar.

You can look at the stats page for an idea of various software that is
submitting spots:
https://pskreporter.info/cgi-bin/pskstats.pl

If someone was remotely accessing an online SDR (Kiwi or not) and
piping that audio to WSJT-X (etc), then it would just be reported as
that specific software with whatever location it was set.

As Sissy has noted, it is common for ops to operate remotely these
days. If an op has a Flex station in Maine but is in Arkansas running
WXJT-X through their remote client, it is up to them to set WSJT-X
correctly. If not set correctly, you would also not be using the
proper grid when making contacts. The same goes for their logging and
confirmations. I've had plenty of POTA stations log me as their home
state, though, so it can be a problem. For that matter, someone
running digital modes physically in a location other than the normal
location could forget to set their WSJT-X to the correct grid and be
reporting spots incorrectly without internet SDR tech.

No matter what, we are depending on someone correctly setting their location.

73,
Luke - AE5AU

> --
> You received this message because you are subscribed to the Google Groups "PSK Reporter" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to psk-reporter...@googlegroups.com.
> To view this discussion visit https://groups.google.com/d/msgid/psk-reporter/31bc35d5-be4c-4fad-9f4b-303ca99246dan%40googlegroups.com.

[Darkly Shining Flotsam]

I'm one such individual (W/SWL/YORK), specifically one using the Web888 which is a variant of the Kiwi effectively. In its case the position it reports is being fed by its onboard GPS fixes, so you rely upon it.

[Sissy KM4QLB]

Good call on DigiSkimmer, Luke.  See screenshot below.

That shows a ham with DigiSkimmer reporting reception of my signal in Switzerland. My ~5W (tr)uSDX rig with attic-mounted, end-fed, random wire antenna is probably not being reliably received nearly 4,200 miles away, especially when you look at the proximity of other spotters of my signal during that brief window.  

Yet that same ham in Swizerland repeatedly reports reception of my signal, day and night, whether conditions are favorable or poor.  He/she has reported reception from more than 800 different stations in the 49.9K reports they made in the last week.  The transmitters are located all over the globe, suggesting to me the the person reporting them is hopping from one remote receiver to another, probably in an automated fashion, leaving us with no idea where our signals were actually received.

(I have redacted their callsign as a courtesy.)

I wish that I could right-click to manually put someone on ignore so that they would be omitted from my map displays.  Maybe in the not-too-distant future, there can be AI analyzing reports for suspicious locations, so that we could have a checkbox to omit reports from people whose grid square locations don't reflect their receiving antenna locations.

-- Sissy KM4QLB

On Thursday, January 30, 2025 at 4:35:11 PM UTC-5 Luke Williams wrote:

[Joni Kähärä]

A major issue with such an "AI" approach would be that there would almost certainly be false positives, and dealing with those would then require further human intervention (if they were even spotted to begin with), and no one is likely willing to do that work. Such a weakness related to automation isn't unique to this ham radio-specific case.

Anyway, is there actual proof instead of gut feeling to show that would indicate that there's something fishy going on here? Switzerland has nice elevation, for example. This doesn't look too weird to me:

https://pskreporter.info/pskmap?preset&callsign=HB9TMC&txrx=rx&band=6000000-8000000&timerange=86400&showtx=1&showgrid=1&showsnr=1&showlines=1&showsnrinlogbook=1&maptype=aea&mapCenter=0,0,1

    Joni OH2EWL

[Philip Gladstone]

It is an interesting problem. If anybody wants to work on it, then I'd be interested in the results. If you go to https://pskreporter.info/csv/2025/02/ then you can download the data that I receive. I'd pick a pair of files like 20250208T080000_e.tsv.gz and 20250208T080000_o.tsv.gz -- these will cover a 6 hour set of observations. [The files both cover the same time period but contain different reports, there are horrible internal reasons why there are two files].

Philip

Philip Gladstone

To view this discussion visit https://groups.google.com/d/msgid/psk-reporter/50635e1c-0b2d-4e09-98e0-ec89a076f03fn%40googlegroups.com.

[Sissy KM4QLB]

> A major issue with such an "AI" approach would be that there would almost certainly 

> be false positives, and dealing with those would then require further human

> intervention (if they were even spotted to begin with), and no one is likely willing to do that work. 

I'd be happy to visit the PSK Reporter website and answer a question about whether my grid square was for my receiving antenna or my butt.  

> Such a weakness related to automation isn't unique to this ham radio-specific case.

It is hard for me to imagine a case in which AI false positives would be less important than in this case.  Using AI as a map filter would presumably be each visitor's choice.  If you can't live with a few false positives, then don't check the box.  

AI, and pseudo-AI, spam filtering generates many false positives, but many people are okay with that tradeoff in order to get less spam.  If you are receiving emails for a suicide prevention hotline, then you probably wouldn't use an AI spam filter.

> Anyway, is there actual proof instead of gut feeling to show that would indicate that there's something fishy going on here? 

Framing it as being either "actual proof" or "gut feeling" is false dichotomy.  I have no 'actual proof' that it will be cold in Helsinki tomorrow, my belief is based on more than a gut feeling.

To be clear, I am not accusing anyone of wrongdoing.  Some of the most popular digital mode software provides no guidance on how to answer the grid square location (i.e., butt or antenna location), so people could come to different conclusions.  It could well be that some, or even most, PSK Reporter users prefer to have butt locations rather than antenna locations.

-- Sissy KM4QLB

[Philip Gladstone]

We want to have people report their antenna locations as that gives other people a better view of the actual propagation. Having said that, the existing consistency checks are based around the distance between the locator and country which issued the callsign. Maybe I have the restrictions incorrect.

Philip

Philip Gladstone

To view this discussion visit https://groups.google.com/d/msgid/psk-reporter/ef5e2fc1-4ebe-4bb4-9e91-e1111af2993cn%40googlegroups.com.

[Sissy KM4QLB]

I want them to report actual antenna position, too.  Perhaps a banner reminding people to make sure that their reported grid sqaure matches their receiving antenna location would be good enough.

Your checks may be fine and the reports I found to be suspicious might be explained by a combination of a superb receiver, a great antenna, very little interference, and lots of elevation.

-- Sissy KM4QLB

[Darkly Shining Flotsam]

Can be hard for Philip to enforce as all too often those posting are doing so via third party tooling vs direct uploads; this maybe something to hit up the providers of those tools (and the likes of the Kiwi crew themselves) to ensure that their settings pages have such a similar banner

[Daryl Popowitch]

Jerk

To view this discussion visit https://groups.google.com/d/msgid/psk-reporter/c33d511f-d5a5-4dab-8bc8-b28c6a5ded2fn%40googlegroups.com.