udp packet delay

209 views
Skip to first unread message

Lily

unread,
Nov 2, 2015, 4:28:39 PM11/2/15
to ns-3-users
Hi all,

In my test, I use udp client to generate traffic and use extract statistics from flowmonitor.  I calculate the average packet delay with (delaysum)/ rxPackets.
I found that when the bandwidth is not saturated the average delay keep same.   I think the delay should increase with the growth of flow traffic rate.
For example, the delay is x , when flow traffic rate is 1Mbps, and then the delay is y, when flow traffic rate is 2Mbps.   x should be greater than y.  But in ns3 , they are same. 
 
Could you give me some hints on how to solve the problem? Or what's the situation in real network?  Thanks.

Tommaso Pecorella

unread,
Nov 2, 2015, 6:54:27 PM11/2/15
to ns-3-users
It heavily depends on the technology you're simulating. In most systems the difference is not noticeable unless you use really advances statistical analysis, and also in these cases it can be hard to notice.
As an example, Wi-Fi tx delay depends only on the back off timers, and the are changing only if a node experiences a collision, which is not that common (unless very close to the saturation point).

T.

Wenrui Ma

unread,
Nov 2, 2015, 7:58:09 PM11/2/15
to ns-3-...@googlegroups.com
One more question, we can set interval and packet size for UDP application. I think it uses constant traffic model.  I want to apply burst model on UDP application.  The average throughput is still same, but in some duration there's no packets and in other duration there's a packet burst. But I have no idea now. Do you have any suggestion? Thanks.


Regards,
Wenrui Ma

--
You received this message because you are subscribed to a topic in the Google Groups "ns-3-users" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/ns-3-users/byDblZ5Weqk/unsubscribe.
To unsubscribe from this group and all its topics, send an email to ns-3-users+...@googlegroups.com.
To post to this group, send email to ns-3-...@googlegroups.com.
Visit this group at http://groups.google.com/group/ns-3-users.
For more options, visit https://groups.google.com/d/optout.

Konstantinos

unread,
Nov 3, 2015, 3:49:15 AM11/3/15
to ns-3-users
Hi,

You can use OnOffApplication to generate UDP traffic with random On and Off times. 
This can mimic bursty arrival process.

Regards,
K.


On Tuesday, November 3, 2015 at 12:58:09 AM UTC, Lily wrote:
One more question, we can set interval and packet size for UDP application. I think it uses constant traffic model.  I want to apply burst model on UDP application.  The average throughput is still same, but in some duration there's no packets and in other duration there's a packet burst. But I have no idea now. Do you have any suggestion? Thanks.


Regards,
Wenrui Ma

On Mon, Nov 2, 2015 at 6:54 PM, Tommaso Pecorella <tomm...@gmail.com> wrote:
It heavily depends on the technology you're simulating. In most systems the difference is not noticeable unless you use really advances statistical analysis, and also in these cases it can be hard to notice.
As an example, Wi-Fi tx delay depends only on the back off timers, and the are changing only if a node experiences a collision, which is not that common (unless very close to the saturation point).

T.


On Monday, November 2, 2015 at 10:28:39 PM UTC+1, Lily wrote:
Hi all,

In my test, I use udp client to generate traffic and use extract statistics from flowmonitor.  I calculate the average packet delay with (delaysum)/ rxPackets.
I found that when the bandwidth is not saturated the average delay keep same.   I think the delay should increase with the growth of flow traffic rate.
For example, the delay is x , when flow traffic rate is 1Mbps, and then the delay is y, when flow traffic rate is 2Mbps.   x should be greater than y.  But in ns3 , they are same. 
 
Could you give me some hints on how to solve the problem? Or what's the situation in real network?  Thanks.

--
You received this message because you are subscribed to a topic in the Google Groups "ns-3-users" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/ns-3-users/byDblZ5Weqk/unsubscribe.
To unsubscribe from this group and all its topics, send an email to ns-3-users+unsubscribe@googlegroups.com.

Wenrui Ma

unread,
Nov 3, 2015, 7:51:59 AM11/3/15
to ns-3-...@googlegroups.com

Thanks. It works.



Regards,
Wenrui Ma

To unsubscribe from this group and all its topics, send an email to ns-3-users+...@googlegroups.com.

To post to this group, send email to ns-3-...@googlegroups.com.
Visit this group at http://groups.google.com/group/ns-3-users.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to a topic in the Google Groups "ns-3-users" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/ns-3-users/byDblZ5Weqk/unsubscribe.
To unsubscribe from this group and all its topics, send an email to ns-3-users+...@googlegroups.com.
Reply all
Reply to author
Forward
0 new messages