I'm trying to check how general and valid protocol suite system ( layered structure ) are.

Does anyone know if radio communication ( broadcast ) via AM or FM , for instance can be viewed in the light of a layered structure ( for example, the OSI model ) as operating at least in one or more in layers ? Maybe could it be fair to say it operates only in the physical layer ?

Now suppose i have a really simple communication system between me and another guy and we implement communication digitally, with error-correcting ( but no signaling like in telephone and also not anything else apart from channel-coding and transmission through the channel) ...

Could it be fair to say that our communication system has only the physical layer ? Or would the communication system to require some protocol for the physical layer ? What would mean a protocol for a really simple physical layer ( like point-to-point one i described in the last example ) ?

I'm just trying to see if we can think about ANY communication system as operating in a layered structure ( like OSI ).


2 Answers 2


Pretty much any communication system can be modelled to some degree within the 7-layer model, though the choice of the layers in the OSI model was clearly aimed at computer networking.

Any system must have a physical layer because communication necessarily requires moving information through space and that is inherently physical. The physical layer deals with the problem of getting a chunk of data (a packet, a byte, whatever) from sender to receiver, not necessarily reliably, and all communication systems must have that.

Beyond the physical layer, things get a bit vague for things that aren't close analogues of computer networking. In your broadcast radio example, you could argue that the error-correcting codes used in digital radio are a basic datalink layer or you could argue that, since there's no mechanism for the receiver to request retransmission of lost packets or to acknowledge receipt, there is no reliable communication so there is no datalink. I'd probably go with the latter: the error correction is nice an' all that but, if a packet's bust, it's bust and you ain't never gonna see it again.

Your second example's a little unclear to me as you don't state whether you're doing anything to establish reliable communication, such as acknowledgments and retransmit requests to ensure that the data gets through. You certainly need some kind of physical layer protocol because you need to agree a meaning for the behaviour of the medium (e.g., this kind of modulation of the radio wave means 0 or this voltage on the wire means 1). If you build anything on top of how to code and decode packets, that's a higher layer in the model.

  • $\begingroup$ Thanks for the answer. How should we treat analog communication ( analog TV or radio broadcast ) regarding errors ? Is there a possibility to minimize errors in an analog transmission, like we do in a digital transmission when we add a channel coder at the transmiter and a channel decoder at the receiver ? $\endgroup$
    – nerdy
    Jul 4, 2014 at 15:00

every communication system is build on layer architechture. beacause each comm. model need data link layer to identify local host and the upper layyer working as well


  • 1
    $\begingroup$ How does this address radio communication in particular? $\endgroup$ Jan 30, 2017 at 8:26
  • $\begingroup$ How does it address two people talking in a room? Remember that the layered models are primarily there to provide terminology for discussing communication systems; systems are not necessarily built according to the layers. $\endgroup$ Jan 30, 2017 at 9:49

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.