6

So I'm having a good bit of trouble understanding this protocol. Many resources seem to compare it to modern networking protocols in some way, while others, including the specification, seem to refer to it as solely a protocol for communicating between the DTE and DCE. I the latter case, I'm extremely confused. How does the DCE get packets to remote terminal at the end of the circuit, since that seems it would involve DCE to DCE communications?

Sorry if this sounds like word soup, as I'm having difficult really explaining my exact confusion. If X.25 is an Interface between Data Terminal Equipment (DTE) and Data Circuit-terminating Equipment (DCE) than how does the data get to the remote terminal specified in the packet? Does the DCE just throw the frames onto the network and it gets switched across various PSE's? How do the PSE's communicate with each other to switch the packets? And what really is a DCE in this case? Modem's? Curiously, I've never seen an X.25 modem even when searching for one (or any X.25 related hardware for that matter, including PAD's or PSE's).

0
11

Technically speaking, X.25 only describes what goes on at the edges -- the DCE/DTE interface, as you guessed. Within the packet switched network itself, any suitable protocol can be implemented. Of course, it has to be able to deliver the signaling and data that are implied by X.25; for example you need to be able to convey a "call request" from source to destination. And data packet boundaries need to be preserved end-to-end (it's a sequenced-packet protocol, not a byte stream protocol).

The mid-to-late 1970s picture looked like this. If you wanted your computer to talk to other computers over a packet-switcched network, then "your computer" would implement the DTE role of X.25.

Some device on Post Office property would implement the DCE role.

In my case, the two would then be connected via a Post Office leased line.

PDP-11 = DTE  
  |  
Synchronous modem   
  | 
Post Office leased line  
  |  
Synchronous modem   
  |  
DCE somewhere in Post Office property  
  |  
+-----------------------------+  
| The packet-switched network |  
|   (the original cloud!)     |  
+-----------------------------+  
  |  
 remote DCE  
  |  
… details omitted …  
  |
 remote DTE   

In this picture, the PDP-11 is equipped with a synchronous line interface, capable of operating in bitstuff mode (e.g., a DUP-11). The modems do what modems traditionally do: condition a digital bit stream for transmission over analog communication wires. There's no X.25 intelligence in them.

(The DTE end had an actual modem as a separate box, with a cable to the line device in the PDP-11; I was not privy to what went on at the DCE end, so that end is a little figurative).

Note that the X.25 packet protocol is a network (layer 3) protocol; to the extent you can map TCP/IP to OSI layers, IP is layer 3, TCP is layer 4. X.25 provides "virtual circuits" that are analogous to real physical circuits.

The X.25 software in "your computer" would probably (certainly in my case) provide some locally-appropriate way for other programs to access X.25 virtual circuits.

A use-case would be DECnet. DECnet networks were (pre-Ethernet) built with point-to-point links, for a fully or partially connected mesh. Any such link could be replaced by an X.25 virtual circuit. Thus DECnet would be the "user" of X.25, and other programs would use DECnet, without knowing that data were travelling over X.25.

5
  • Well put. Except, there are a few different configurations, especially when the (X.25) DTE isn't a generic computer but itself a closed box simply offering one or more ports to the X.25 network – Raffzahn Jan 2 at 22:58
  • The above diagram was what we implemented at DEC (well, we also had X.29 virtual terminal support). Doing something useful with it was up to the customers :-) – another-dave Jan 2 at 23:53
  • Fun part, the test network I got in storage was setup by Nixdorf using DEC VAX machines :)) – Raffzahn Jan 3 at 0:54
  • Where do you live, @another-dave? In the US, I would have expected the DCE to be a Bell minion rather than Post Office. – chrylis -cautiouslyoptimistic- Jan 3 at 6:04
  • Back then, I lived in the UK. – another-dave Jan 3 at 13:56
5

X.25 is anything but an interface. It is a packet-switched network, very much comparable to TCP/IP. Most of the packet-switching, however, is entirely invisible to the end subscriber as PSE (Packet-switch-exchange, comparable to a router in a TCP/IP network) are entirely hidden within the Telco's network - The end user sees a pretty simple serial line (RS-232, 449 or X.21), nothing else. This connection entirely looks like a point-to-point link.

The physical serial interfaces can also be established over modem lines - the modem, however, just forms the lowest physical layer, transporting the end-users physical serial bitstream to the nearest PSE or PAD. These PADs (Packet Assembler/Disassembler) packetise serial data into X.25 packets that can be switched through the network, and are also responsible to establish/tear down the virtual links through the packet network.

1
  • 2
    The title of recommendation X.25 is Interface between Data Terminal Equipment (DTE) and Data Circuit-terminating Equipment (DCE) for terminals operating in the packet mode and connected to public data networks by dedicated circuit -- i.e., X.25 is an interface standard, though colloquially we'd call the PSN an "X.25 network". – another-dave Jan 2 at 21:29
0

The DCE/DTE connection is V.24 aka RS-232. X.25 is a signalling protocol.

0

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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