My first post on this site
– please be gentle!
I am a telecoms engineer, nothing to do with aviation.
Pings – let’s see if we can sort this out.
The satellite is always transmitting. Think of it as a steady “tick” from above. The receiver, not the ACARS system, or any other similar higher function in the plane, just the receiver is always listening to the tick. IN fact the receivers on all the planes and boats are listening to the "tick". By clever encoding the “tick” carries information to say “this is the start of my sequence of messages” at defined intervals, and then it sends data intended for specific receivers in various time slots. Of course, all receivers "hear" all the messages, but they only actually “listen” to the ones intended for them. But the background “tick” keeps going all the time, so every receiver can synchronise itself all the time, keeping track of where the sequence of messages starts, just listening and waiting for a message from them. When a receiver recognises a message is for that particular receiver (i.e. plane), of course it “listens” to the message and acts accordingly, passing the message on to the appropriate unit where necessary. But one of these messages it watches for is a message saying “are you still there”. When it gets that message the receiver has to reply in the correct “slot” to say “still here”. So now the sat system on the plane has to transmit to reply. In this case the reply is effectively an automatic response from within the satellite receiver – nothing else needs to be involved.
But notice the response to the “are you still there” message has to be sent in the correct time slot. Remember the receiver is listening to the “ticks” from the satellite all the time – not just when there is data to exchange. All the time, so the plane has a “clock” ticking at the same rate as the satellite’s clock. But, the plane’s clock is running late all the time as it takes time for the “tick” sent out by the satellite to travel the 35,000 km to the receiver. The equipment on the plane then has to reply in the correct slot after it has received the “are you there” message and so it replies. Now, the satellite has a problem. Because it does not know how far the plane is away, it actually does not know when the reply to its message is going to arrive. It cannot do what the plane has done and keep a “copy” of the plane’s clock ticking away because, firstly, the plane is not transmitting its own “tick” for the satellite to follow, and in any case it would need to have a similar clock running for every plane and ship using the service. In any case it does not need to.
What it does is allow a window for the return message to arrive. It has a petty good idea since it knows the earth is 35,000 Km away as a minimum, less the altitude of a plane, of course. It knows that the “receiver” has to respond so many ticks after receipt of the request. It knows the receiver has a good clock tick signal – it is the very signal the satellite is transmitting. So it has a time window when a response can be expected. The reply will arrive at the early end of the window if the message is from a transmitter directly under the satellite, later from one at the edge of the coverage, earlier from a plane than from a ship etc.
Now, here is the bit I have not seen mentioned before. The “tick” is related to the frequency of the satellite’s radio band. It has little to do with the relatively slow rate at which data can be sent over the communications channel. The frequency of Inmarsat C is around 1600 MHz. That is 1,600,000,000 “ticks” a second. So the timing resolution the satellite can see is to an accuracy of 1/1600000000 of a second. Light is fast. But the amazing speed of electronics we now have is such that we can measure time with astonishing precision. Just think about those laser tape measures you can buy. You can measure a distance to a millimeter, and they work by timing pulses effectively, all in something you can buy for a few £/$/Eu. Think about a GPS set. An ordinary GPS set will give you a position with in a few metres, limited not by the ability of measure time, but by various vagaries in the transmission of the signals. A “pro” surveying GPS set does clever things to reduce the vagaries, and while it may take a few minutes to give a good “fix” they can give your position to typically 15mm accuracy. Given longer to analyse and process the signals, weeks in some cases, to an accuracy of considerably less than 1mm. Note, accuracy, not resolution. Absolute accuracy, and all without a fantastically expensive atomic clock at the “user end” though there are several in
each orbiting GPS satellite.
So, back to our satellite. It was expecting a response at “tick” number XYZ and it actually arrives at “tick” number XYZ + whatever. It knows, because it is in the standard, that the receiver on the plane is obliged to reply exactly N ticks after receipt of the request, so by subtracting N from XYZ + whatever it knows how long the message has taken to transit to the receiver and back again in “ticks”. Halve that and you have how many ticks away the plane is. One “tick” is 1/1600000000 of a second. In a vacuum- and remember part of this journey is not in a vacuum – 3 x 10 to the eight metres. So, divide this by 1600000000 i.e. by 1.6 times 10 to the nine and you get about half a metre as the potential accuracy (the satellite will have a VERY accurate clock onboard) and resolution of ths measurement. In principle the satellite could determine how far away the responding receiver/transmitter was to a precision in the order of a metre. In practice there would be no point in measuring this accurately and in any case the presence of the atmosphere and variations in the precise timing in the receiver makes this rather “optimistic”. These are, after all, communications systems not navigation systems. The satellite does not need to know to this level anyway, it can work with signals arriving in the allocated window.
This is a very simplified explanation, not always desperately accurate, but basically shows how it works. For example the frequency used for the plane to reply is actually slightly lower, but it will be locked to the satellite’s transmission clock tick.
As a further explanation, think of a lighthouse flashing away constantly with a constant beat. It flashes white. You sit and watch holding a torch. You watch and watch and tap your foot to the beat of the white light. You are told that if the light flashes red you MUST flash your torch back three beats later. And you do exactly that, exactly three beats later. The man at the lighthouse is watching and from his point of view he sees a torch flash back not three beats later, but three and a bit beats later. The three beats if the time you have been allowed to wake up, get your finger on the flash button, the “and a bit” is the time it has taken the light to get from his lighthouse to you plus the time it has taken your torch light to get back. Clearly in this case you could not measure the distance in this way, but you can see the principle.
As has already been shown, the satellites are at a very precisely known height over the earth, so equal distances from the satellite are where concentric cones intersect with the earth – circles on the earth’s surface representing equal elevations of the satellite.
My suspicion is, and I say no more than a suspicion, that the satellite’s system can also measure the frequency of the reply accurately and so have some estimate of the relative velocity of the plane relative to the satellite, but the geometry will limit how much use this would be. Clearly the plane’s altitude when transmitting is relevant; you would not be able to differentiate between a low plane nearer the point the satellite sits over and a higher plane further away.
I also suspect that the satellite would keep a log noting that it had received a response the “are you there” messages at such and such times, but it would only keep the technical details of the latest transaction, as this would help in allocating its time slots efficiently. Remember, while memory to hold this sort of data is cheap, on a satellite power is the limiting factor all the time, and weight adds to the “delivery” cost of getting these things up there. So the commercial birds would not want to retain data that they did not need to hold for longer than necessary. The logs may be downloaded to the ground too, but even downloading takes power.
Hope that helps. I have been reading all the posts since this thread opened. There are some on here that simply waste effort – and show complete disrespect, but there are some incredibly insightful and helpful posts here conveying real information. Not all are “hard” facts, but many are fascinating, like the description of what it is really like in a search plane. Many thanks to the posters.