p2p link - queues

71 views
Skip to first unread message

H D

unread,
Mar 16, 2024, 4:37:18 PM3/16/24
to ns-3-users
Hello, given a p2p link between 2 nodes, how can I define their queues and the scheduling policy between the queues for each node?

H D

unread,
Mar 16, 2024, 8:05:16 PM3/16/24
to ns-3-users
note: i am not sure how i can use this API - https://www.nsnam.org/docs/models/html/prio.html  to implement a scheduling policy like Round-Robin, Weighted-Round-Robin, Strict-Priority and specifically, what is the meaning of the priomap in this context. i would appreciate some help. Thanks :) 

ב-יום שבת, 16 במרץ 2024 בשעה 22:37:18 UTC+2, H D כתב/ה:

Tommaso Pecorella

unread,
Mar 17, 2024, 6:39:31 AM3/17/24
to ns-3-users
P2P links are - essentially - what the name says: point to point links: two unidirectional wires paired to make a bidirectional link. The devices do not contend for the channel, as each device has its own "wire".
Hence, there's no need for a inter-device scheduling system, each device do its own.

What I'm trying to explain is that that Tx queue system in P2P devices prioritizes only the packets being sent by that particular device, and it doesn't care about the packets of other devices - because it doesn't need to.

Lastly, before you ask: no, changing NetDevice won't work (not as you might think). The decision of what packet to send first is made by the Tx queue of each device, but the decision of which device have to send something is the role of the MAC layer. The overall result MIGHT be similar, but it's far more complex than simply choosing a queuing policy. 

H D

unread,
Mar 17, 2024, 6:55:59 AM3/17/24
to ns-3-users
I'm sorry, but it seems like my question wasn't clear enough.
I meant to ask how I can set up queue policies for a specific router.
For example, how can I define 4 output queues with a Round Robin scheduling policy for it?

ב-יום ראשון, 17 במרץ 2024 בשעה 12:39:31 UTC+2, Tommaso Pecorella כתב/ה:

Vikram Dubey

unread,
Mar 17, 2024, 7:00:03 AM3/17/24
to ns-3-...@googlegroups.com
can anyone send me the code with awk script file of TCP Reno , NewReno And Tahoe.

--
Posting to this group should follow these guidelines https://www.nsnam.org/wiki/Ns-3-users-guidelines-for-posting
---
You received this message because you are subscribed to the Google Groups "ns-3-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ns-3-users+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/ns-3-users/90a36fce-ad6b-4213-90c9-8b81424031cen%40googlegroups.com.

Tommaso Pecorella

unread,
Mar 17, 2024, 7:48:29 AM3/17/24
to ns-3-users
Each output port does have its own scheduling policy (and its own queue).

A (single) output queue can have different policies to handle its own packets.

As far as I know (but I might be wrong), there's no queue that fits your needs that is (also) usable by P2P - mainly because each NetDevice has its own ways to mark packets, so each NetDevice defines its own subclasses of Queue.

Said so, it shouldn't be difficult to create a new Queue subclass that does what you ask, the only things you'll have to re-define is the behaviour of Enqueue and Dequeue according to some info in the packet you enqueue.
Check out "WifiMacQueue", it does something similar (redefine a queue). In your case it might be slightly more complex, as the packet queued in a router doesn't carry the info of the source port - only the info about the originator. However, it's definitely possible to do something useful, perhaps by extending the P2P class.

Vikram Dubey

unread,
Mar 17, 2024, 7:59:06 AM3/17/24
to ns-3-...@googlegroups.com
can anyone send me the code with awk script file of TCP Reno , NewReno And Tahoe.

--
Posting to this group should follow these guidelines https://www.nsnam.org/wiki/Ns-3-users-guidelines-for-posting
---
You received this message because you are subscribed to the Google Groups "ns-3-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ns-3-users+...@googlegroups.com.

Tommaso Pecorella

unread,
Mar 17, 2024, 9:27:34 AM3/17/24
to ns-3-users
@vikram: please read the positing guidelines.

This is the very last warning. Failing (again) will result in harsh measures.

Last but not least, since this is the last warning and the next violation will result in [you can imagine], let me explain something simple.

First: your request doesn't say what you did (or attempted to do) to solve your problems.
Second: you don't specify the scenario you're interested to, so any reply will be (almost for sure) not right.
Third: you can't expect that someone does YOUR job for you.
Fourth: your request shows that you didn't even bother to check what ns-3 can or can not do, and if there are examples that might actually help you.

Shortly put - you're:
1) Spamming
2) Not asking for help, suggestions, or clarifications
3) Trying to lure someone to do your job

Said so, since I'm harsh, but also willing to give you a last chance, if you'll show to have understood your mistakes you **might** receive some replies.
If not... how's the weather in Hyderabad? Cloudy for what I can see.
Reply all
Reply to author
Forward
Message has been deleted
0 new messages