Announcement

Collapse
No announcement yet.

Selection of Radar sites for web display?

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #46
    The problem is that you continue to try to use whether your radar ID appears on the map or not to decide whether your data feed is "good" or "bad" when the logic behind whether your radar ID appears or not is so obfuscated by the number of developers who have made changes to the algorithm that it's not usable to draw any conclusion on whether your data feed is "good" or "bad".

    Comment


    • #47
      Try using playback for the last 24 hours at 40x or 50x and filter your radar. You might be surprised at how many times you see your callsign.
      F-KDAG1

      Comment


      • #48
        I did it. Got two groups of four and a few separate single for last 24 hours... Now I see other four. Probably a grand total of less than 20?
        "Aircraft seen" shows 2094 for the last 24 hours. That is not what I would call "statistical average".

        Unless... there are really 1000 receivers in the 150-180 miles radius around me.
        Last edited by SoNic67; 2018-09-30, 23:12.

        Comment


        • #49
          Originally posted by SoNic67 View Post
          But i guess you can't understand that, even after I have stated it many times. And I am not sure why you care about this topic, nobody forces you to read it. Or post for that matter.
          After all, everything was probably "discussed" already, so we should shut down this forum altogether?
          I admit it. I don't understand why you keep labouring your point. But that doesn't stop me responding to your troll like posts.

          Now you're questioning the frequency of your label appearing and getting a proportionate "display time". Unfortunately it still boils down to the original answers, as provided by Mike (who runs FR24) and others in this and other threads.

          The radar label is meaningless. It's best ignored. Let Mike and the rest of the FR24 team manage the problem, if one exists. You can't say they haven't noticed your posts.
          Mike


          www.radarspotting.com

          Radarspotting since 2005

          Comment


          • #50
            Originally posted by SoNic67 View Post
            I can't fathom that, sorry.
            Local 97, web zero... that makes no sense, even with your explanation. What, there are more than 100 receivers around me? Even then I should see sometimes the planes tracked by my site. I never see any.
            First of all. You see 97 ADS-B and MLAT aircraft. MLAT aircraft will never show a feeder ID, so you have to exclude them from the calculation.
            A majority of aircraft in US are tracked with MLAT, so you are probably only seeing 25-45 ADS-B aircraft.
            And yes if you feed from the US East Coast, there are probably 100+ feeders tracking the flights going at crusing altitude over your house.

            Comment


            • #51
              Originally posted by Mike View Post
              First of all. You see 97 ADS-B and MLAT aircraft. MLAT aircraft will never show a feeder ID, so you have to exclude them from the calculation.
              A majority of aircraft in US are tracked with MLAT, so you are probably only seeing 25-45 ADS-B aircraft.
              And yes if you feed from the US East Coast, there are probably 100+ feeders tracking the flights going at crusing altitude over your house.
              Thanks for being on-topic, it's a good point about MLAT. I wasn't aware that the "Aircraft Seen" on website includes that. Right now I have on the "Aircraft Uploaded" 70 planes from a total of "Aircraft Tracked" of 85, I guess that I can adjust the number with that approximate ratio.
              And yes, I don't know exactly the number of local sites that I am connected in MLAT on FR24, but on FA it shows "synchronized with 140 nearby receivers", so probably that's a correct estimation.
              I guess I can look on some of my neighboring sites, to see their 24 hour stats, but I have a feeling it's not 1:1000 like looks to be in my case.

              Originally posted by Anmer View Post
              But that doesn't stop me responding to your troll like posts.
              The only troll on this thread is you. You are the one that keep saying that you are bored by this thread and therefore everyone should stop posting. You are the one that requires me to shut up my own thread. Typical "generation me" attitude.
              If you don't have anything constructive, on topic, to say, just... don't. You are not the only person here on forum, or in the world.
              Last edited by SoNic67; 2018-10-07, 11:38.

              Comment


              • #52
                Originally posted by SoNic67 View Post
                The only troll on this thread is you. You are the one that keep saying that you are bored by this thread and therefore everyone should stop posting. You are the one that requires me to shut up my own thread. Typical "generation me" attitude.
                If you don't have anything constructive, on topic, to say, just... don't. You are not the only person here on forum, or in the world.
                I don't recall using the word "bored" or suggesting everyone should stop posting. Or that you should "shut up". But that's precisely what you're telling me to do.

                And guess what? I think you're right. I may not be the only person in the world. What a shame.

                But if it makes you feel better, so be it.
                Mike


                www.radarspotting.com

                Radarspotting since 2005

                Comment


                • #53
                  Originally posted by SoNic67 View Post
                  Does anyone know what are the criteria based on which the planes are associated with radar sites on the web page?
                  For example, my radar site picks up a plane that took of from an airport close-by. I look on FR24 live tracker and it's there, and down, where that radar dish is shown it says Radar "T-KPHF1" and shows my site. Good.
                  The plane flies away, I can still see it on my local web-page (ADS-B), but, at not even 5 miles from me, the "Radar" changes to F-KFKN1. Well, that's a station at least 30-40 miles away, and probably it just picked up that plane. Why was it switched from my feed?
                  Why I ask this? Because most of the time, when I filter the Live tracker page I see zero planes as being associated to my radar. But I see multiple associated with other sites (mostly F-KFKN1, F-7NC21).

                  My question is... why? Is something that makes my data un-trusty or un-reliable? Because then, I would like to fix that.

                  Is that based on the maximum range? One of the sites doesn't share it's stats, the other indeed has a bigger range. I was imagining that the hand-over happens when I drop the signal, another radar is selected (from the many that are seeing that plane). Not that is preemptively switched at 10 miles out.
                  I would like to revisit SoNic67's initial post, and respond to it in the spirit and context originally intended.
                  I too have wondered why we see the results displayed on the web site the way they do, and why one Radar is often selected over others over a period of time.


                  My comment isn't about whether a radar code should be displayed or why do I care anyway, it is just about SoNic67's initial question.


                  I have read Mike's comments and I applaud Mike for contributing to it the way he has. In reality he has described an extremely complex process which has been tweaked and enhanced over quite a few years to deliver the results it does today in such a timely way.


                  I too have gone through all of the steps that SoNic67 has probably been through, largely making physical changes to antenna placements, lengths of cables, efficiency of network, speed of internet service, and trying combinations of different antenna types, multiple DVBTs and the odd Beast. So like SoNic67 I have a (software) engineering background although not as up to speed as Oblivian is in his knowledge of the software aspects of Linux.

                  I did consider that speed of the internet feed may be affecting whether an individual's data feed is being selected for a particular data element used in the web display. But after having gone from ADSL to Fibre 100 I observed no perceptible change to my level of profile, and as all feeders in a particular area gradually move to new service oferings, such apects probasbly diminish.


                  I think what Mike implied is true, that geographic location will be significant as well as your setup so that particular radio waves will behave in a local manner. I have noticed that one of my radars is occasionaly attached to flights from an aero club located on the other side of the city, across which many other commercial flights are transitioning.


                  Putting that to one side, I would expect that optimisation processes have been introduced over the years to the FR24 server software to ensure that things run as quickly as possible.
                  Often historical patterns can often become embedded in the processes, through various lookup indexes and the like so that certain feeder's data will be utilised over others.

                  Mike has indicated that there is no deliberate ranking of feeders, the data selection is random, and so we can infer that there is no deliberate selection of radars for a priority list.

                  However I would suggest that based, on my own and clearly many others observations there is some pattern occurring regarding the display of Radar codes against planes.
                  My view as I have indicated is that the anomolies noticed are caused by the way the server software is operating, as they are utilising historical patterns for the selection and processing of data, as part of it's optimisation logic. Call it AI if you like. :-)
                  .

                  Comment


                  • #54
                    Originally posted by JohnSunnyhills View Post
                    Mike has indicated that there is no deliberate ranking of feeders, the data selection is random, and so we can infer that there is no deliberate selection of radars for a priority list.

                    However I would suggest that based, on my own and clearly many others observations there is some pattern occurring regarding the display of Radar codes against planes.
                    My view as I have indicated is that the anomolies noticed are caused by the way the server software is operating, as they are utilising historical patterns for the selection and processing of data, as part of it's optimization logic. Call it AI if you like. :-)
                    .
                    Thank you, that makes so much sense! I can accept that, even if probably is not 100% accurate.

                    Comment


                    • #55
                      For example, right now F-7NC21 has at least 30 planes shown, 20 of them clearly closer of my radar. I have none shown from my radar (T-KPHF1).
                      Granted he has a 288nm max range and my radar only 198nm. But statistically... looks like that radar is always preferred by the software (cached?).
                      Last edited by SoNic67; 2018-10-23, 14:31.

                      Comment


                      • #56
                        It used to be the case that in a contest ,to acquire the same aircraft, between a F feeder (FR24 supplied) and a T feeder (every other type), the F feeder would always win because it is faster (does certain operations in hardware that the others do in software). I have not seen this topic discussed for a long time in the forums probably because no one takes much notice of feeder codes any more. Things may well have changed but I suspect not because around most big airports there seem to be a handful of F feeders and a much larger number of T feeders - more than a hundred in some cases. It is quite conspicuous how many aircraft are attributed to the F feeders compared to the T feeders.

                        Comment

                        Working...
                        X