Take the 2-minute tour ×
Super User is a question and answer site for computer enthusiasts and power users. It's 100% free, no registration required.

I understand that a network card detects collisions because their signal has a higher amplitude than normal signals. But why is that signal of higher amplitude?

For example, say we have the simplest of cases, two computers communicating half-duplex on one wire and they both transmit at the same time, a collision occurring:

A |------>-<------| B

Why does the current intensity arrive at the stations at a higher level? The currents are opposed one to each other.

Most probably I'm missing basic physics here, but I could sure use an explanation...


share|improve this question

migrated from stackoverflow.com Dec 2 '10 at 5:13

This question came from our site for professional and enthusiast programmers.

The network cable is a transmission line and signals are attenuated as they propagate along the transmission line. However collision detection is a little more complex than just measuring amplitudes. –  Paul R Nov 30 '10 at 10:50
I imagine it's more complex then that, but I can't find nowhere how it really works. Do you have some links I could look into? All i found until now is just how the protocol works, but nothing at the physical level. –  Claudiu Nov 30 '10 at 11:08
Well CSMA/CD is pretty ancient these days - I think it was only used on coax cable Ethernet and maybe on early twisted pair networks, because these were half duplex. Modern networks are typically full duplex and don't use CSMA/CD, AIUI. –  Paul R Nov 30 '10 at 13:17
I know it's ancient, I know all about the inner-workings of the protocol, the thing I don't know is how stuff happens on the physical level...and so far besides lectures on how the protocol works (things you can find on every wiki page) and notices that the method is too old for me to be interested in it, I didn't manage to get a decent answer :) –  Claudiu Nov 30 '10 at 22:19
This question probably can be answer by math. You should see Stack Exchange for Math –  SgtOJ Dec 2 '10 at 5:19

1 Answer 1

up vote 2 down vote accepted

I might be a bit late in answering this, but nevertheless it might help those who reach here in the future.

Quoting from this page:

When a NIC has data to transmit, the NIC first listens to the cable (using a transceiver) to see if a carrier (signal) is being transmitted by another node.

When there is data waiting to be sent, each transmitting NIC also monitors its own transmission. If it observes a collision (excess current above what it is generating, i.e. > 24 mA for coaxial Ethernet), it stops transmission immediately and instead transmits a 32-bit jam sequence.

This patent seems to be describing a mechanism to detect collisions. Skimming through it, I believe it works by analysing the bits which the sender recieves. A phase decoder sits on the sender and it has a 2-bit output. Two of the outputs are reserved for 1/0 and remaining 2 indicate a collision has occured.

share|improve this answer
Very detailed, I was looking for something exactly like this. Thanks! –  Claudiu Feb 24 '14 at 8:10

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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