Distributed algorithms such as leader election problems(LCR and HS algorithms)

[karthik kondapally]

Hello,

I am quite new to NS3 , and i have read the documentation.
 
i am trying to simulate a Distributed algorithm, LCR (a leader election algorithm in ring topology (uni directional)) . i have read the Mpi module, but  could not understand ,how to use it .

algorithm requires a Traffic generator at each node (in ring topology) which genearates a single packet(containing a rank or id) every round . But in MPI only node can generate traffic.

if i have missed any module that can help in implement this  algorithm, please guide me to that module.


Thanks and Regards,
karthik.

[Biyu Zhou]

2013/11/26 karthik kondapally <konda...@gmail.com>

Hello,

I am quite new to NS3 , and i have read the documentation.
 
i am trying to simulate a Distributed algorithm, LCR (a leader election algorithm in ring topology (uni directional)) . i have read the Mpi module, but  could not understand ,how to use it .

algorithm requires a Traffic generator at each node (in ring topology) which genearates a single packet(containing a rank or id) every round . But in MPI only node can generate traffic.

i guess it will be this:  MPI will generate multiple LPs, and each LP (represents a node) will run the same script, so in MPI simulation, every node will generate traffic.

 

if i have missed any module that can help in implement this  algorithm, please guide me to that module.


Thanks and Regards,
karthik.

--
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 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/groups/opt_out.

[pdbarnes]

Actually it's not like that at all.

For the original question, MPI is orthogonal to the model you discuss, so forget about MPI for now.  Just define the ring topology, and an application (or pair of client/server applications) to generate/receive packets and handle your LCR algorithm.

The MPI module let's you split a large model (many nodes) across many processes (MPI ranks) to get higher execution rate of the model, compared to the default serial implementation.  Generally you want to get your model working at modest scale first, then tackle MPI parallelization.

Peter

On Tuesday, November 26, 2013 6:55:12 AM UTC-8, Biyu Zhou wrote:

2013/11/26 karthik kondapally <konda...@gmail.com>