must a spectrum model be global?

[bjfons...@gmail.com]

Hi,

As I learn how to use the SpectrumModel of NS3, I've noticed that the spectrum models of wifi, lte, and lr-wpan all use global definitions for the spectrum model.

Must a SpectrumModel be global? is there a particular reason for having a spectrumModel be global?

B. Fonseca

[Tommaso Pecorella]

Hi,

yes, it must be global to be able to calculate the interferences between different sub-systems.

It would be nice to be able to split the instances at run-time (it would save a LOT of computational power), but it would mean to forecast who's going to interfere with who in all the simulation. This, in turns, depends on the nodes positions and the channels they're using... a huge problem.

Anyway, for now it's a single global.

T.

[bjfons...@gmail.com]

Hi Tommaso,

Sorry for the late reply.

Thanks for the explanation.

B.Fonseca