Solved: T 1 history and background

sarahr202 · ‎02-25-2009

Hi every body!

i was reading about the history behind T1 line. I am reading the book" Network warrior" by o'reilly press.

This is what my book says" In the 1950s, the only method for connecting phone lines was with a pair of copper wires. For each phone line entering a building,there had to be a pair of copper wires. In 1961, Bell Labs in new jersey invented the T1 as a means for digitally trunking multiple voice channels together ."

Based on above,a small town of 200 homes in 1950, which require phone connectivity,we would have needed 200 pairs of copper wire to connect them to CO in 1950. Is it correct to assume?

Book also mentions " smart jack" where the digital circuit T1 terminates. Since T1 on client side terminates on csu, i assume ' smart jack" terminates T1 at CO. Am i correct ?

Thanks a lot!

badalam_nt · ‎02-25-2009

Look at:

http://www.sundance-communications.com/cgi-bin/ultimatebb.cgi?ubb=get_topic;f=17;t=000032;p=0

http://books.google.ro/books?id=n4p4WLJXWc8C&pg=PA13&lpg=PA13&dq=smart+jack+T1&source=bl&ots=ZRsU2b0x3D&sig=asIrLd7jvCNPNltJHsZe0gN_nqA&hl=ro&ei=58WlSYvtPNKX_ga0v4WbBQ&sa=X&oi=book_result&resnum=3&ct=result#PPA13,M1

(page 13)

http://www.tomshardware.co.uk/forum/page-21765_17_0.html

Each phone line was carrying one phone call, the original frequency bandwidth of an analog voice call was between 300Hz and 3400Hz (so only the medium frequencies were carried, no Hi-Fi sound supported)

Then digital telephony came into the market.

The analog voice signal was sampled each 125 microseconds, then each sample was converted to digital (12 bits per sample were generated) and finally there was a compression of each sample from 12 bits to 8 bits. It means finally you got 1 byte each 125 us, so the rate of a digital voice signal is 64 kbits/s (also known as DS0 in North America).

This is what the audio codec ITU G711 miu law is doing (in Europe it is ITU G711 A law, the difference is the compression algorithm used for going from 12 bits to 8 bits).

By multiplexing 24 such lines of 64kbits/s each you get the 1536kbits, that is T1.

The DS1 signal takes twenty-four of these DS0 channels and multiplexes them into a single bit stream,

giving a rate of 1.536 Mbps.

(note that only 23 useful phone lines could be used, as you need to allow signalling as well inside the T1).

So yes, each house had its own pair of copper, carrying one voice conversation only, and the pair was going from teh client till the service provider.

But there were methods to reduce the number of pairs or wires, by installing concentrators close to clients neighbourhood, connecting for instance 50 clients to this concentrator box, and then from the concentrator till the voice provided with only 12 pairs of wires. They used the statistic that at busy hours normally max 12 lines will be used at the same time. Of course it introduces "blocking", as the 13th will not be able to call, but that is the compromise to cost (12 lines cost less than 50 lines). And then came this multiplexing which improved the cost, without trading against "overbooking".

View solution in original post

badalam_nt · ‎02-25-2009

Look at:

http://www.sundance-communications.com/cgi-bin/ultimatebb.cgi?ubb=get_topic;f=17;t=000032;p=0

http://books.google.ro/books?id=n4p4WLJXWc8C&pg=PA13&lpg=PA13&dq=smart+jack+T1&source=bl&ots=ZRsU2b0x3D&sig=asIrLd7jvCNPNltJHsZe0gN_nqA&hl=ro&ei=58WlSYvtPNKX_ga0v4WbBQ&sa=X&oi=book_result&resnum=3&ct=result#PPA13,M1

(page 13)

http://www.tomshardware.co.uk/forum/page-21765_17_0.html

Each phone line was carrying one phone call, the original frequency bandwidth of an analog voice call was between 300Hz and 3400Hz (so only the medium frequencies were carried, no Hi-Fi sound supported)

Then digital telephony came into the market.

The analog voice signal was sampled each 125 microseconds, then each sample was converted to digital (12 bits per sample were generated) and finally there was a compression of each sample from 12 bits to 8 bits. It means finally you got 1 byte each 125 us, so the rate of a digital voice signal is 64 kbits/s (also known as DS0 in North America).

This is what the audio codec ITU G711 miu law is doing (in Europe it is ITU G711 A law, the difference is the compression algorithm used for going from 12 bits to 8 bits).

By multiplexing 24 such lines of 64kbits/s each you get the 1536kbits, that is T1.

The DS1 signal takes twenty-four of these DS0 channels and multiplexes them into a single bit stream,

giving a rate of 1.536 Mbps.

(note that only 23 useful phone lines could be used, as you need to allow signalling as well inside the T1).

So yes, each house had its own pair of copper, carrying one voice conversation only, and the pair was going from teh client till the service provider.

But there were methods to reduce the number of pairs or wires, by installing concentrators close to clients neighbourhood, connecting for instance 50 clients to this concentrator box, and then from the concentrator till the voice provided with only 12 pairs of wires. They used the statistic that at busy hours normally max 12 lines will be used at the same time. Of course it introduces "blocking", as the 13th will not be able to call, but that is the compromise to cost (12 lines cost less than 50 lines). And then came this multiplexing which improved the cost, without trading against "overbooking".