NS3 is there, but a lot of researchers are sticking with NS2.
I know that ongoing studies that started basing on NS2 will continue with it, but why new studies don't use NS3?
Is NS3 not stable enough yet?
Anyone who tried it and found bugs in some models please feel free to tell.
The main reasons for people not to move are:
1) ns2 is well known and it does have a lot of models. Some models haven't been re-developed for ns-3, and some will never be.
2) lazyness, who knows ns2 thinks that he/she can stick to it, no matter the bugs.
3) lazyness (2), ns-3 is really different from ns2, so it's like re-learning a lot of things.
The main reasons to *move* to ns-3 are:
1) ns2 is a dead horse. Its core developers aren't maintaining it anymore except for very small changes. All the development moved to ns-3.
2) ns-3 is faster, easier, more stable, supported.
3) ns-3 has a lot more features than ns2, like... well, check it out.
It's like asking why people still use Fortan when there's C++ around. Who knows Fortran like their hands is free to use it, but I wouldn't suggest for a new user, unless really needed.
Is it true that NS2-based publications are no longer accepted in international journal? I found some papers based on NS2!! How ???
This is pure bulls**t. Any serious journal will never reject a paper because it uses a specific tool.
However, using the wrong tool can cause a paper rejection.
ns2 is old, mostly not maintained, and not actively developed. Said so, some ns2 models can be still useful, especially if they're not present in more recent simulators (like ns-3).
ns-3 is definitely more advanced and more versatile, and if I'd have to choose between ns2 and ns-3, I'd definitely go for ns-3.
- Badges
- Science topic
- Similar topics
- Computer Communications (Networks)
- Simulators