Question on T1 and TDM

[Yohhan]

When someone talks about different channels and time division multiplexing on a T1 carrier, are we talking about a single physical cable? What kind of cable is used for T1? I'm doing some reading for class, and it's difficult to tell sometimes if they're talking about something physical or logical.

 

[AFB]

TDM=A static set transfer rate( I think ) Its not a different cable. I think they are refering to a logical portion of the total capacity for that cable.

 

[bgroff]

Originally posted by: Yohhan
When someone talks about different channels and time division multiplexing on a T1 carrier, are we talking about a single physical cable? What kind of cable is used for T1? I'm doing some reading for class, and it's difficult to tell sometimes if they're talking about something physical or logical.

If you're doing this reading for class, then I assume you must be familar with the 7 layer OSI model by now. The T-carrier structure can be looked at in respect to the OSI model. On the physical layer, you have the cabling. For a T-1, the telco usually delivers an RJ-48c (8 pins, looks like ethernet) "smartjack" that is connected to the customer's CSU/DSU. On the data access layer, you have the channels. For a T-1, there are 24 DS0 channels that make up a DS-1 (T-1). This is carried over a single cable from the telco to the customer.

Does that make any more sense?

 

[Yohhan]

Yes, that makes sense. So when someone talks about a "channel" in general, it's referring to layer 2?

 

[ScottMac]

Actually, no.

TDM stands for Time Division Multiplexing: basically slicing a chunk of time into twenty-four timeslots (the frame actually contains some signaling and error-checking as well). Each slice of time represents one "channel. " As mentioned previously, each channel is a DS0 - "Dee-Ess-Zero" - equivelent to one phone line or 64K of data (possibly 56K, depending on the system). A basic T1 is strictly a Layer One system (essentially it is just "a wire" ... even though it is using some signaling to actually move the traffic).

When you apply an additional signal to the "wire" ( the T1), it will be in the form of HDLC, SDLC, Frame Relay, X.25 / X.21 ... whatever - THEY are the Layer Two Signaling (the "Frame" containing the layer three info (which contains the layer four info which contains the layer five info ... etc)).

The pulses that represent the basic T1 signaling (the pulses representing the basic multiplexed information) would be comparable to the pulses seen on a cable carrying Ethernet. Ethernet is a Layer Two protocol (so is Token Ring, so is Arcnet ...etc). The PULSES / signaling carrying the Ethernet frame would be comparable to the pulses / signaling carrying the SDLC/HDLC/Frame-Relay layer two Frame.

T1 is carried over two pair (one Transmit pair, one Receive pair, pins 1&2 and 4&5 on an RJ45). The spec calls for "premises cable," which is individually shielded twisted pairs inside an overall shielded sheath / jacketing. Most T1 CPE (Customer Premises Equipment) is built to tolerate UTP (Cat - Anything), and it will work ... it's just not the spec.

The SmartJack is usually the demarcation point where the Telco's responsibility ends. It is a device that has some surge protection and logic or mechanism to allow the Telco to "loop" for diagnostic purposes (connect the Tx and Rx such that a signal sent to the SmartJack is returned back to the source). Some Telcos will do testing to the CSU/DSU, but they are usually only obligated to test to the Smartjack.

If there is an additional span of cabling from the SmartJack to the CSU, it is considered an "extended demarc" and may or may not be considered for testing (i.e., "loop the CSU and test").


To get back to the original point, do not confuse the term "frame" when discussing the "Super Frame" (SF) or "Extended Super Frame" (ESF). That frame (SF or ESF) is pure layer one as opposed to the Frame used in HDLC / SDLC/ FR/ etc which is the layer two protocol.

Hope that helps ...


FWIW

Scott

 

[bgroff]

If you look at the T-carrier system from the perspective of a broader networking system, then it most certainly is layer 1. If you are looking at the T-carrier system in and of itself without respect to higher networking protocols, it can be viewed as a structured networking system. Of course, not all seven layers are clearly defined, but the same can be said for other networking protocols/stacks. The T-carrier system is more like a 3 layer system... There's the electrical transmission layer (it isn't necessarily always going over copper), there's the data link layer (voice or data, SF or ESF, AMI or B8ZS), and there's a nebulous networking layer (troubleshooting, other SF/ESF messages). Each layer is independant of the other.

Bet you never looked at it like that before, huh?