Super User is a question and answer site for computer enthusiasts and power users. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I found that the distance that data transmission can be done in Ethernet has to be reduced when the transmission speed increases, which in turn leads to the use of Carrier Extension, Frame bursting, etc. Is noise a fact that determines this distance? Then how does noise affect this transmission distance?

share|improve this question
up vote 1 down vote accepted

If I'm understanding your question correctly, its probably not noise (interference from external EM/Power sources), but Signal Attenuation (degradation of the signal traveling through the wires, caused by the imperfect nature of transmitting signal on copper). this would lead to more corrupted or lost frames, which in the case of TCP data, would trigger resends.

in the old days of bus networks, collisions were an issue, and CSMA/CD was employed to avoid otherwise mitigate collissions caused by two machines starting to transmit at approximately the same time, such that for the cable length, the sending machines had to pause for the length of time it took a 64k signal to traverse the max cable length at wirespeed, causing a relationship between transmission rate and cable length based on the logical operation of the network (not the physical).

share|improve this answer
64k as in 64 karacters? – Hennes May 14 '13 at 15:35
Thanks @frank. But I found from a book that a sender station has to keep on sending data to a receiver station to maintain an essential property that sender is still transmitting when the noise burst gets back. I still couldn't clearly understand what's the reason for that. – Deepal Jayasekara May 14 '13 at 15:35
@Hennes, you are right, I meant bytes. as you get older, things that were big become small, and things that were small become forgotten.... thanks for the reminder. – Frank Thomas May 14 '13 at 15:43
@DeepalJayasekara, IIRC (its been a lot of years), the reason for that is that until the signal has propagated to the end of the bus cable, and come back, you cannot be sure that a collission has not occured on the far end of the bus. that means that at for a given wirespeed and cable length, the time to traverse teh whole cable must be approximately the same amount of time that it takes the sending nic to put 64B on the wire. all things being equal, that means that as wirespeed increases, the cable length decreases, unless the capacity of the cabling can be increased as is the case with UTP. – Frank Thomas May 14 '13 at 16:32
more info – Frank Thomas May 14 '13 at 16:40

Both 10-baseT and 100-baseTX have a 100 meter specification. However, to reliably get 100m out of 100-baseTX, the cable has to conform to ANSI(TIA/EIA-568) specifications, particularly the use of either the T568A or T5678B (see previous hyperlink) wiring pinout standards. Each unshielded twisted pair of wires in cat5 have differing rates of twists per meter, to reduce crosstalk and interference. 10-baseT, however, only requires cat3 cable, which has a less rigourous specification, so in the case where a 100m cable is not terminated to spec, has been bent too sharply (less than 4x the outside diameter of the cable), or is otherwise degraded, it may no longer be fit to run 100-baseTX, but still adequate to run 10-baseT. However, when cables conform to the specifications above, both 100-baseTX and 10-baseT are good to 100 m.

share|improve this answer
Thank you @Nevin! – Deepal Jayasekara May 16 '13 at 8:58

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .