How jamming signal works in CSMA/CD?

Assume station A sends a frame. When this frame almost reaches station D, D also sends a fame. A and D are at the two far ends of a same segment.

A

D
10101010101010101010101010101010>x<1010

When D detects the collision (much earlier than A), D will send out jamming signal. Since A has not detected the collision, and continues to transmit, will D's jamming signal collide with the rest of A's frame?

A

D
000000000000101010110101011010><-jamming signal

If the jamming signal collides with the remanent bits, will other nodes recognize it as jamming signal at all?

Find more posts tagged with

Free for TechExams community: Cybersecurity salary guide

Compare cert salaries and plan your next career move

Button

Comments

Bellman80

I thought it was always the sending station that does the jamming signal. The 100m max segment length and frame size has to do with being able to detect a collision and send a jamming signal while the frame is still leaving an interface.

johnifanx98

Bellman80 wrote: »

I thought it was always the sending station that does the jamming signal. The 100m max segment length and frame size has to do with being able to detect a collision and send a jamming signal while the frame is still leaving an interface.

Both A and D sends frames. Just D sent a bit later than A, thus the collision occurs closer to D.

EdTheLad

So you are looking at a half-duplex setup, lets say 10BaseT or 100BaseT. 2 devices connected back to back A and D, with twisted pair. The cat5 cable between uses 2 pairs for transmit and 2 pairs for receive. Wires 1,2 are transmit, wires 3,6 are receive in respect to A, the jamming signal from D cant collide with whats being sent from A as its on a different wire

johnifanx98

EdTheLad wrote: »

So you are looking at a half-duplex setup, lets say 10BaseT or 100BaseT. 2 devices connected back to back A and D, with twisted pair. The cat5 cable between uses 2 pairs for transmit and 2 pairs for receive. Wires 1,2 are transmit, wires 3,6 are receive in respect to A, the jamming signal from D cant collide with whats being sent from A as its on a different wire .

To be more accurate, I'm looking at half-duplex on coaxial cable instead of hub.

I knew nobody would use cable or even hub, but just figure the idea of this contention protocol might still be useful...

EdTheLad

With coax, there are still two cables, tx and rx, so it's exactly the same behavior as copper.

johnifanx98

EdTheLad wrote: »

With coax, there are still two cables, tx and rx, so it's exactly the same behavior as copper.

Oh, please. I've ever setup a LAN myself using coaxial. it's a single cable with one single copper core.

EdTheLad

Yes your right, i never worked with it so i just assumed it was two as per my experience working with E1 which uses 2 coax. Anyway in that case, back to original question, the jamming signal is used to keep the line busy so all other devices in the collision domain reset their wait timers. So those access bits that collide with the jamming signal just create noise on the line for a finite time until its a pure jamming signal. The end result is the same, other devices see the line busy and reset their wait timers.