link budget of satellite communication?

How to calculate the carrier to interference ratio (C / I) in link budget of satellite communication?

  • Where in the chain do you wish to know ? - the levels vary as the signal progresses through the RX.

    Noise and interference are two different things, and need to be calculated separately.- an out of band interferer is (hopefully) not amplified as much as the wanted signals, assuming the designer has their filtering strategy correct. Noise (in band) once accompanying the signal cannot be removed. In a well-designed system the RX noise level is set by the front end devices, as after that the signal level is high enough that the noise of successive stages has only a small additional effect.

    The other part is the link budget of course is the level of  the wanted signal. How much is transmitted and what (very small) fraction of that is intercepted by the RX antenna..

    Satellite links are one of the few cases when an unobstructed path can be assumed and the inverse square law actually applies - so you need the altitude and the TX power.

    Antena 'gain', or more accurately "directivity" can be converted to an equivalent capture area.

  • Thank you.

  • I really don't think I can add much (if anything)  to what mapj1 has said. I am also sure that you have a good understanding and knowledge of this field.

    As you appreciate, satellite link budgets are simple and you don't need any specialist tools.  When I worked in satellite communications, we used spreadsheets for the link budgets, but at university we calculated them by hand on paper. One of the best known textbooks on satellite communication systems was authored by Maral and Bousquet, and I believe that this covers satellite link budgets.

    I imagine that you are probably interested in the overall link budget giving you the signal degradation and hence bit error rate (ber)* at the ground-based receiver.  
    As mapj1 said, propagation over satcomm links is line-of-sight (LoS)* so the signal is limited by noise rather than interferers. I.e. it is the Carrier to Noise Ratio (CNR)* rather than C/I ratio that dominates - unlike multipath propagation in e.g. cellular mobile communications, where C/I dominates.

    I have not heard of co-channel interferers causing degradation in civilian geostationary satellite communication systems. (It is easy to accidentally transmit on somebody else's uplink channel, thereby causing interference until you realise the mistake and select the correct frequency.)  For ordinary geo-stationary satellite communications, the separation of the satellites' orbital slots and the necessary directivity of the uplink antennas prevent uplink interference with other satellites that share the same frequencies. Similarly, interference from other geostationary satellites' downlinks is prevented by the directivity of the ground terminal's antenna. I have not worked on inter-satellite links yet.

    Of course, it's possible that you have a satcomm system in which ground terminals are perhaps handheld mobile terminals, similar to phones (e.g. satellite phones) that are used like other mobile phones in crowded urban areas amongst tall buildings etc that might cause local multipath scattering of the LoS signal to/from the satellite. I have not worked with this type of ground system and don't know if it is likely that you would receive interference from other mobile terminals of the same system: I expect that their uplink transmissions should not interfere with downlink channels received by mobile terminals of similar systems. I don't know if this describes the use of mobile terminals for such systems as e.g. Thurya, Inmarsat or Iridium when used in urban locations on land.

    My experience was with geostationary satellites and their ground systems.  Perhaps your system does not use a geostationary satellite. If your system utilises a constellation of satellites in e.g. Low Earth Orbits with multiple satellites in view at the same time and crossing the sky at high speed, perhaps C/I might be a consideration. However, I imagine that the engineering of such systems would take care of such C/I problems by design as a problem of handover/hand-off. In that case, I imagine the C/I calculations should be published.

    I suppose that if your satcom system does suffer from interference, you will need to add the interferer signals to the wanted signal at the point where they enter the link and at the power at which they join the link.

    I apologise for the length of this response.

    * I follow the old IET/IEE practice of spelling out each abbreviation the first time I use it.

  • I am not sure if the OP(original poster) is about any more, as there have been no recent posts by that username.

    In my reply, I was not too sure of the correct level of complexity, and decided to opt for 'keep it simple' - but your additional comments do add useful extra information.

    My own  experience is more with mobile things in green or sand-coloured boxes, and then interference from the other side can sometimes be very much  a thing to be reckoned with, both on channel and on adjacent frequencies. Even for non -comms systems, like GPS, it is sometimes the case that more directional antennas are used to check that the signal that is supposed to be coming from a satellite is coming from the right sector of the sky and following the right track. Ground-based signals are treated with great suspicion, although knowing their bearing is still sometimes useful... So the control of antenna side-lobe responses is something that probably gets more attention than the civilian equivalent,

    As you suspect the hand held satellite phones are indeed vulnerable to local multipath, as the antennas have to be pretty much omni-directional and while the link margin has some spare tens of dBs to allow for that, even so they  really work best with a good view of the sky - but this is not such an issue, as normally in a built up area there is usually an alternative local network. Thuraya you mention is actually  not very much different under the bonnet to a slowed down GSM having both up-link and  down-link frequency separation, and TX RX timeslots. There is the advantage that the tx- rx distance (and therefore the signal level/dynamic range ) is not as variable as a terrestrial network - no user is ever really near the satellite after all!

    It is worth noting that Mr Musks 'starlink' system is an exception, and instead solves this ground clutter issue  another way, and the ground units do orient the antenna to track the satellites, and the flat panel array can be configured by combining and phasing the internal antenna elements to give either a wide beamwidth for searching or a narrower one for tracking. There are also  an eye-watering number of satellites compared to any previous system, but this is matched by the high no of paying customers and the high available bandwidth, and it ensures pretty much unbroken connection.


  • A helpful answer...thank you.

  • That is great answer...thank you Mike.

  • Hello Mike,

    Thanks for the reply and for filling in the gaps that I had been guessing at. I should have looked those things up before mentioning them.

    I thought your reply said all that could sensibly be said at this point and that it was pitched at exactly the right level - unlike my clumsy contribution. I'm not sure why I replied, particularly since the OP must have acted on your helpful response long ago. 

    I must admit ignorance of the Starlink system: it's something that I have not looked into.

    The great majority of my satcomms experience was with the space segment, initially for satellites where interferer problems such as you mentioned were very important. I appreciate the usefulness of identifying the bearing of an unwanted interferer so that you can mitigate appropriately.

    My Mobile Satellite Service payload experience was on an Inmarsat project rather than Thuraya. I cannot remember too much about the air interface of the Inmarsat traffic, I vaguely seem to remember thinking that the channelisation, had some similarity with GSM inasmuch as I think it was defined by both timeslot and frequency. (I don't know if we even described it as an 'air interface'.)
    I didn't know whether Thuraya or Iridium used Frequency Division Duplex at the ground terminal but fully expected that they did just like Inmarsat 4 and every Fixed Satellite Service or Broadcast Satellite Service that I have ever heard of. I should have realised that satellite mobile phones use terrestrial networks when available. 

    I left the satellite company to work on satcomm ground systems of various sizes from subsystems in rugged boxes to immobile ground stations with large dishes for various customers before I moved to another field many years ago.

    Best regards,

  • Hello Mr Mekasser,

    It is my pleasure. I am glad if my answer was any help. 

    Best regards,