Adhoc collisions?

41 views
Skip to first unread message

jamp...@gmail.com

unread,
Mar 6, 2017, 9:53:54 AM3/6/17
to ns-3-users
I created ~8000 wifi nodes in a square space (with a limited trasmission range) and the simulation works in this way:
So the messages will go from the center to the borders. There are no waiting times between the messages, so there should be a lot of collisions. But when I execute the simulation, I see that all the node will receive the message and will do a broadcast. Why? Could it be that there are so many nodes that the message will reach everyone? Is there a way to see when and where there are collisions?

Tommaso Pecorella

unread,
Mar 6, 2017, 8:05:49 PM3/6/17
to ns-3-users
Hi,

you can look at the MacLow, but keep in mind that Wi-Fi is really good at avoiding (and resolving) collisions, provided that the channel is not completely saturated.

T.

mida...@gmail.com

unread,
Mar 7, 2017, 10:34:40 AM3/7/17
to ns-3-users
I'm interested in this problem as well because I'm working on a similar thing. From what I understand from the post, there are hundreds of wireless nodes sending broadcast messages at the same time (and there should be several collisions in that example) and as you just said, Wi-Fi is good at managing collisions. What could be a way to create the same simulation and get collisions? In the first link I see that the user gives "WIFI_PHY_STANDARD_80211b" to the WiFiHelper. Could changing that standard be enough?
Reply all
Reply to author
Forward
0 new messages