So you want your station to decode frames so you can climb up the dB leaderboard and catch @csete and @DL4PD and my station 146 riyadh. A nice game, as competition between people always brings out the best in them.
The Fox birds and XW-2 birds cw beacons are a good start, all on VHF.
Unisat-6, over its ground station in UK, also transmit lots of frames. Bugsat and falconsat on UHF gives many frames as well. Lituanicasat-2 and IRAZU also gives double figures frames decode on certain passes. Please feel free to add more recommendations!
I wonder though, how else can we calculate this? If everyone started focusing on the high framerate satellites, would we neglect the others? What is the value in 1,000 CAS-4A packets versus one WOD packet from another satellite in the same time frame?..
I don’t have ideas/answers here, nor do I want to discourage the leaderboard, just want to keep the conversation going.
Loosely related, now that we are able to decode telemetry into a tsdb we can easily see the rate of packets getting decoded by day. Maybe there is something to be said for a station that contributes during a low point in the data? (see UWE-4 for example below)
Quality is always better thean quantity. At the end it the whole idea is to get your station to decode a satellite, not just track it.
What about giving more weight to difficult to decode satellites or giving leaderboard for each satellite? It would be nice if a static station can decode satellites whne not in use?
Anyway lets track and decode as many birds as possible
in light of QO-100 being decoded / decodable I wanted to revive this topic (as it would be trivial for someone to setup a constant receiver and climb/stay at the top of the leaderboard, not to mention fill the db)
setting aside the current technical limitations of doing this in db today (as there are many):
Should we limit the leaderboard to orbiting sats?
What would an algorithm look like for “quality over quantity”? The higher percentage of frames in db for a sat, the lower the modifier on contributions from a station?
I like the “leaderboard for each sat” idea, filed it here.
I like the idea of a leaderboard for observations! Each observation has (at least) the following characteristics which we can base a score:
Duration - I believe we want to promote longer observation durations since they utilize the network better. A linear or exponential function could be used to contribute to the score. E.g.
Decoded status - The decoded status should definitely contribute to the score. E.g.
Maximum elevation - Successful lower passes indicate a better station. It should also be included to the formula. E.g.
Age of transmitter - We want to promote reception of newer satellite and transmitters. Thus we can use again some exponential function to do that. E.g.
Frequency - Observations on higher frequencies are more difficult. We should promote that as well. E.g.
Satellite success rate - We also want to promote observations on rare satellites. The score should depend on good versus bad past obs. E.g.
Observation status (through vetting input) - If the obs status is bad, the score should definitely be zeroed.
The overall score with the addition of weight then becomes:
Not always the case if, say, a station has a very limiting horizon. (such as the vertical yagi test) - this scenario also goes against the “Maximum elevation” modifier.
So, I’m not sure if we should make assumptions about difficulty, as one station may be more successful at certain parameters than another, and visa versa (especially when comparing, say, one fixed station against another).??
Yeah, I am thinking about the future. Sooner or later all observation artifacts will be ending up in DB where they belong.
The duration parameter has to be there somewhere, even in the form of “actual station utilization”. Otherwise it won’t be fair. Take for instance 2 stations which both do 20 observations a day. They should not have the same score when one has on average 10min obs and the other 1m obs.
hmm I want to agree, yet there’s an aspect of “quality over quantity”. 2 stations return 1 packet each, one station gets a higher score for that packet because it was picked out of a 10 min obs vs the other station which grabbed it in 1 min. The latter station could really be the “better” at receiving and decoding, and grabbing that data in shorter time means it may be freed up for another observation elsewhere (which doesn’t match my vertical yagi / narrow horizon use case), yet would be penalized.
So yeah, like you say “station utilization” might be the modifier here rather than a duration modifier specific to the observation.
Thanks for the input, I like where this topic is going.
@cshields, we could then use a linear function for the duration. Assuming that the station score is the sum of observation scores, then actual utilization is simply duration over leaderboard time range.
In other words, equal sum of short and long obs will get the same score.
Moving forward to auto-scheduling I don’t think it matters so much who schedules an observation.
However giving a shout-out to the observer in the community forum or in social media, if something special was caught like a non-alive satellite or an unscheduled event is not a bad idea.
Just to make it clear… I don’t mean that scheduling isn’t an important task, just that I’m not sure if we need to add a point system/credit system on it with the auto-scheduling coming.
