Relation between delay and no. of nodes

47 views
Skip to first unread message

SEDE

unread,
Dec 29, 2021, 2:12:59 AM12/29/21
to OMNeT++ Users

Hello, I am experimenting INET MAC protocols by changing the no. of nodes. If I increase the number of nodes in the network, with X-MAC, the delay per received packet decreases. Would anyone suggest me why the delay decreases in this case? Thank you.

Alfonso Ariza Quintana

unread,
Dec 31, 2021, 11:41:28 AM12/31/21
to omn...@googlegroups.com
Have you checked the path size?
There is the possibility that if you increment the number of nodes, the mean path length could decrease.

De: omn...@googlegroups.com <omn...@googlegroups.com> en nombre de SEDE <afroz...@gmail.com>
Enviado: miércoles, 29 de diciembre de 2021 8:12
Para: OMNeT++ Users <omn...@googlegroups.com>
Asunto: [Omnetpp-l] Relation between delay and no. of nodes
 

Hello, I am experimenting INET MAC protocols by changing the no. of nodes. If I increase the number of nodes in the network, with X-MAC, the delay per received packet decreases. Would anyone suggest me why the delay decreases in this case? Thank you.

--
You received this message because you are subscribed to the Google Groups "OMNeT++ Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to omnetpp+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/omnetpp/d56abf50-eac5-45e3-9885-98f5e0b411b9n%40googlegroups.com.

SEDE

unread,
Jan 2, 2022, 3:32:52 AMJan 2
to OMNeT++ Users
Thanks for you feedback, Alfonso. How I can check the path size?

With the topology where the nodes are static, why the mean path length could decrease because of increasing the number of nodes?

Would you please elaborate it a bit more? Thank you.

Alfonso Ariza Quintana

unread,
Jan 3, 2022, 11:39:08 AMJan 3
to omn...@googlegroups.com
You can use the TTL parameter, smaller ttl values implies more hops, you can record the mean ttl value for the received packets in the application module and later compute the mean ttl of all nodes.

Enviado: domingo, 2 de enero de 2022 9:32
Para: OMNeT++ Users <omn...@googlegroups.com>
Asunto: Re: [Omnetpp-l] Relation between delay and no. of nodes
 

SEDE

unread,
Jan 3, 2022, 7:19:24 PMJan 3
to OMNeT++ Users
Sir, what are the possible reasons of the mean path length could decrease because of increasing the number of nodes? Would you please suggest? Thank you.

Alfonso Ariza Quintana

unread,
Jan 4, 2022, 4:36:44 AMJan 4
to omn...@googlegroups.com
Probability, if there are more nodes, the probability to select a closer destination node can increase

Enviado: martes, 4 de enero de 2022 1:19

SEDE

unread,
Jan 4, 2022, 4:43:29 AMJan 4
to OMNeT++ Users
Sir, I am considering a single-hop topology where there is only one destination. In that case, what can be the reasons of getting mean delay per received packet decrease with the increase of the number of nodes? Thank you Sir.

Alfonso Ariza Quintana

unread,
Jan 4, 2022, 6:28:35 AMJan 4
to omn...@googlegroups.com
X-MAC with one hop,
In this case, the results should be similar, the problem is that you need to repeat the simulation with other seeds,
if you don't repeat multiple times, there is the possibility that with more nodes the delay could be smaller because you can have more probability that the destination node could be near to awake. In X-MAC more of the delay is the time that it is necessary to wait for the wake-up of the destination

Enviado: martes, 4 de enero de 2022 10:43

SEDE

unread,
Jan 4, 2022, 9:55:24 PMJan 4
to OMNeT++ Users
Sir, I already considered 5 repeatation and got such results. Would you please suggest how many times I should repeat to get the expected results? Many thanks.

Alfonso Ariza Quintana

unread,
Jan 6, 2022, 5:22:31 AMJan 6
to omn...@googlegroups.com
If you are using a protocol of type anymac, more neighbors decrease the delay, but if you are using unicast, and the destination can be any of the neighbors, the mean delay should be similar in a long simulation, in short simulations the results could be no converge.
And you need to check the variance and confidence interval, it is possible that the results could be overlapped.




Enviado: miércoles, 5 de enero de 2022 3:55

SEDE

unread,
Jan 6, 2022, 6:17:42 AMJan 6
to OMNeT++ Users
Sir, I am using X-Mac. The topology is all nodes are sending packets to 
only one destination (similar to the inet sensornetwork topology). In such case, 
if the number of nodes increases, delay per packet decreases. Did you mean this should 
be the expected result? 
I could not understand what does 'anymac' mean. Thank you.

SEDE

unread,
Apr 23, 2022, 6:14:27 AMApr 23
to OMNeT++ Users
Hello Alfonso, I have tried with multiple repetition but still the result is same. Would you please suggest on my above query? Thank you.

"I am using X-Mac. The topology is all nodes are sending packets to 
only one destination (similar to the inet sensornetwork topology) located at the center. In such case, 
if the number of nodes increases, delay per packet decreases. Did you mean this should 
be the expected result? 
"

Alfonso Ariza Quintana

unread,
Apr 23, 2022, 3:00:13 PMApr 23
to omn...@googlegroups.com

I am not sure, it is possible that could be due to the statistics, more nodes reduce the between the node start to transmit and the time that the node awake

 

Enviado desde Correo para Windows

 

De: SEDE
Enviado: sábado, 23 de abril de 2022 12:14
Para: OMNeT++ Users
Asunto: Re: [Omnetpp-l] Relation between delay and no. of nodes

 

Hello Alfonso, I have tried with multiple repetition but still the result is same. Would you please suggest on my above query? Thank you.

 

"I am using X-Mac. The topology is all nodes are sending packets to 

only one destination (similar to the inet sensornetwork topology) located at the center. In such case, 

if the number of nodes increases, delay per packet decreases. Did you mean this should 

be the expected result? 

"

 

On Thursday, January 6, 2022 at 10:17:42 PM UTC+11 SEDE wrote:

Sir, I am using X-Mac. The topology is all nodes are sending packets to 

only one destination (similar to the inet sensornetwork topology). In such case, 

if the number of nodes increases, delay per packet decreases. Did you mean this should 

be the expected result? 

I could not understand what does 'anymac' mean. Thank you.

 

On Thursday, January 6, 2022 at 9:22:31 PM UTC+11 Alfonso Ariza Quintana wrote:

If you are using a protocol of type anymac, more neighbors decrease the delay, but if you are using unicast, and the destination can be any of the neighbors, the mean delay should be similar in a long simulation, in short simulations the results could be no converge.

And you need to check the variance and confidence interval, it is possible that the results could be overlapped.

 

 

 

Enviado: miércoles, 5 de enero de 2022 3:55


Para: OMNeT++ Users <omn...@googlegroups.com>
Asunto: Re: [Omnetpp-l] Relation between delay and no. of nodes

Sir, I already considered 5 repeatation and got such results. Would you please suggest how many times I should repeat to get the expected results? Many thanks.

On Tuesday, January 4, 2022 at 10:28:35 PM UTC+11 Alfonso Ariza Quintana wrote:

X-MAC with one hop,

In this case, the results should be similar, the problem is that you need to repeat the simulation with other seeds,

if you don't repeat multiple times, there is the possibility that with more nodes the delay could be smaller because you can have more probability that the destination node could be near to awake. In X-MAC more of the delay is the time that it is necessary to wait for the wake-up of the destination

Enviado: martes, 4 de enero de 2022 10:43


Para: OMNeT++ Users <omn...@googlegroups.com>
Asunto: Re: [Omnetpp-l] Relation between delay and no. of nodes

Sir, I am considering a single-hop topology where there is only one destination. In that case, what can be the reasons of getting mean delay per received packet decrease with the increase of the number of nodes? Thank you Sir.

On Tuesday, January 4, 2022 at 8:36:44 PM UTC+11 Alfonso Ariza Quintana wrote:

Probability, if there are more nodes, the probability to select a closer destination node can increase

Enviado: martes, 4 de enero de 2022 1:19


Para: OMNeT++ Users <omn...@googlegroups.com>
Asunto: Re: [Omnetpp-l] Relation between delay and no. of nodes

Sir, what are the possible reasons of the mean path length could decrease because of increasing the number of nodes? Would you please suggest? Thank you.

On Tuesday, January 4, 2022 at 3:39:08 AM UTC+11 Alfonso Ariza Quintana wrote:

You can use the TTL parameter, smaller ttl values implies more hops, you can record the mean ttl value for the received packets in the application module and later compute the mean ttl of all nodes.

Enviado: domingo, 2 de enero de 2022 9:32
Para: OMNeT++ Users <omn...@googlegroups.com>
Asunto: Re: [Omnetpp-l] Relation between delay and no. of nodes

 

Thanks for you feedback, Alfonso. How I can check the path size?

 

With the topology where the nodes are static, why the mean path length could decrease because of increasing the number of nodes?

 

Would you please elaborate it a bit more? Thank you.

 

On Saturday, January 1, 2022 at 3:41:28 AM UTC+11 Alfonso Ariza Quintana wrote:

Have you checked the path size?

There is the possibility that if you increment the number of nodes, the mean path length could decrease.

Reply all
Reply to author
Forward
0 new messages