Notes to Chapter 3, Sections 3.3 - 3.8

page 63   DS system
Do not worry about eventually having to memorize all the different system capacities.  At test time, you will have to know that a DS-0 "building block" channel has a data rate of 64 kbit/sec, and that you can calculate the data rates of larger, multiplexed systems by multiplying the number of channels being used.  Thus, if you know that a DS-1 uses 24 DS-0 channels, you should be able to calculate 24 x 64,000 = 1.536 Mb/s

The information you are paying to send goes by various names, but "payload" is probably the most common one.  Overhead is not of interest to the sender and receiver, and they never see it, but it takes up digital "space."  To make an analogy with a physical product, the product's packaging is a necessary overhead cost.  You can't have the product without the box it comes in, and you'd rather not pay for it, but you have to.

page 70   T1 carriers
What's the difference between a DS-1 and a T1?  DS-1 is a signaling system, and a T1 is a carrier (physical connection) that uses a DS system.  The European E-  and Japanese J- carriers are based upon the DS system.  (See the web link for specifics.)

page 71    T1 pulse train
Note that in this system you will never see two positive pulses in a row, or two or more negative pulses in a row.  If the last 1 was a positive pulse, the next 1 will be a negative pulse.  It would seem easier to use zero volts for a 0 and positive volts for a 1, the way it is done inside a computer.  Why go to all this trouble?  It turns out that there are less errors this way, and the average of all the positive and negative pulses is zero, which is useful for transmitting DC voltage on the line.

page 72    B8ZS
The combination of positive-going and negative-going pulses works great when there is data on the line, but what happens when a string of all zeroes is transmitted?  The plan doesn't work any more.  To solve this, B8ZS was developed.  Don't worry about the details of B8ZS, just know that it gets used when a string of 8 consecutive zeroes is put on the line.  The name should now make more sense:  Bipolar (positive or negative choice) with 8 Zero Substitution.

page 74    frame relay
Telephone traffic is "circuit switched."  If you are talking with someone in California, there is a complete physical connection between you and your friend, and this connection is open (and you are paying for it) whether or not you are talking or there is a pause in the conversation.

"Packet-switching" is designed for the bursty nature of data transmission, where there are long pauses that no one wants to pay for.  In packet-switching, the data is divided up into many small pieces, which are sent individually from the sender and then reassembled at the receiving end.  The individual packets may take different routes, and may take varying times, to get to their destination, depending upon how busy or available the different routes are.  Packet switching is the basis of the Internet and its TCP/IP suite of protocols.  Greatly simplified, IP is the addressing scheme, and TCP is the disassembly/reassembly/error-checking scheme.

Frame relay uses "virtual circuits" to transmit the packets of data.  A virtual circuit is not a continuous physical connection, as with a telephone connection, but it is a continuous circuit as far as the user can tell.  In reality, different circuits are used as their availabilities go up and down.

page 75    structure of the frame relay frame
The frame header includes the number of the DLCI ( pronounced "dell-see"), which keeps track of the sending and receiving locations.  The FCS is the Frame Check Sum, which is used to check for errors.  The checksum is calculated from the data at the sending site, and then recalculated from the received data.  If the two match, there were no errors in the transmission.

page 77    ATM
Every 48 bytes of data sent via ATM requires a 5-byte header.  The payload percentage is thus 48/(48+5), or 48/53, or 90.6%, with an overhead of 9.4%.