Fernando Ariel Gont wrote in a message to Ismail Al-Sulby:
FAG> But I don't understand when you say that after noticing that
FAG> the carrior is free, you send your packet AND _*keep*_
FAG> _*sensing*_ _*the*_ _*carrior*_ ...
Let's say that each card puts a 1 volt DC bias on its signal. Then, by
measuring the average DC bias, you can deduce how many cards are transmitting
simultaneously. If a transmitting card sees the DC bias rise higher than it
should, then it knows that some other card is also transmitting and a
collision has occurred. Because it takes real time for the signal to travel
along a cable, it can easily happen that two cards decide to transmit within
a couple of bit times of each other, and a collision results.
FAG> What would happen if:
FAG> Two machines were waiting for the carrior to get free.
FAG> The line becomes free, and both machines notice it at the
FAG> same time. They both try to send their packets, but at the
FAG> same time!
Right. They use the DC bias measurement to detect this. The frame is lost
anyway, but the time the cable is unavailable is minimized.
-- Mike
---
---------------
* Origin: N1BEE BBS +1 401 944 8498 V.34/V.FC/V.32bis/HST16.8 (1:323/107)
|