ns2 Vs ns3

[Mukul Shukla]

I have just started to use ns3 for my research and teaching. I have
some background of ns2.My area of research is Security in Mobile Ad
Hoc Networks. I want to clarify the following doubts regarding ns2 and
ns3 before choosing one for the research.

1. I have read the difference between ns2 and ns3 from the tutorial of
ns3. I want to know which simulator will be useful for my research
area?
2. I have seen that ns-2.34 was released way back in 2009. Will there
be any new release of ns2 or the new development will always take
place in ns-3.
3. Are there any plan's to maintain the ns2 and it's documentation or
it is advisable to shift to ns3 for new projects? As the documentation
and books of ns2 are very thorough, I wish more tutorials, course work
and books should be published for ns3.
4. Some protocols of ns2 are yet to be written for the ns3 (eg, DSR
and Dymo routing protocols). Are there any plans to develop a
Developer's Manual, similar to ns2 Manual for coding the new
protocols?
5. Will there be any visualiser as NAM in ns2?
6. Is the integration of ns3 on IDE like Eclipse or Netbeans planned?

Thanking you.

[Lalith Suresh]

Hello,

On Mon, Jul 11, 2011 at 11:12 AM, Mukul Shukla <mukulr...@gmail.com> wrote:

I have just started to use ns3 for my research and teaching. I have
some background of ns2.My area of research is Security in Mobile Ad
Hoc Networks. I want to clarify the following doubts regarding ns2 and
ns3 before choosing one for the research.
 

1. I have read the difference between ns2 and ns3 from the tutorial of
ns3. I want to know which simulator will be useful for my research
area?

I've used it for research on routing protocol security in OLSR and it's good. :) I've never used ns-2 except for some university assignments, so can't comment on that.

Regarding ns-3's limitations itself for security research, one limitation I'm aware of is that we don't model processing delays within nodes, so you might have issues simulating something like a CPU DoS on a node.

 

2. I have seen that ns-2.34 was released way back in 2009. Will there
be any new release of ns2 or the new development will always take
place in ns-3.

3. Are there any plan's to maintain the ns2 and it's documentation or
it is advisable to shift to ns3 for new projects? As the documentation
and books of ns2 are very thorough, I wish more tutorials, course work
and books should be published for ns3.

Both projects aren't exactly related (ns-3 is a complete re-write from scratch), but I can surely say that more development effort is going into ns-3 than ns-2, and ns-2 is only maintained rather loosely.

 

 

4. Some protocols of ns2 are yet to be written for the ns3 (eg, DSR
and Dymo routing protocols). Are there any plans to develop a
Developer's Manual, similar to ns2 Manual for coding the new
protocols?

What exactly would you like to see in such a manual? To develop routing protocols for ns-3, have a look at src/olsr, src/aodv or src/dsdv to see what you need to do. In essence, you'll need to create a class that inherits from Ipv4RoutingProtocol or Ipv6RoutingProtocol and extend it with the functionalities you require.

For DSR, there's been an announcement here: http://mailman.isi.edu/pipermail/ns-developers/2010-July/008203.html

 

5. Will there be any visualiser as NAM in ns2?

There is PyViz, and NetAnim.

 

6. Is the integration of ns3 on IDE like Eclipse or Netbeans planned?

If you dig around in this mailing list, you can find people who've used Eclipse to work on ns-3.

 

Thanking you.

--
You received this message because you are subscribed to the Google Groups "ns-3-users" group.
To post to this group, send email to ns-3-...@googlegroups.com.
To unsubscribe from this group, send email to ns-3-users+...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/ns-3-users?hl=en.

--
Lalith Suresh

Department of Computer Science and Engineering

Instituto Superior Técnico
www.lalith.in

[Tom Henderson]

On Jul 11, 7:45 am, Lalith Suresh <suresh.lal...@gmail.com> wrote:
> Hello,
>

> On Mon, Jul 11, 2011 at 11:12 AM, Mukul Shukla <mukulrshu...@gmail.com>wrote:
>
> > I have just started to use ns3 for my research and teaching. I have
> > some background of ns2.My area of research is Security in Mobile Ad
> > Hoc Networks. I want to clarify the following doubts regarding ns2 and
> > ns3 before choosing one for the research.
>
> 1. I have read the difference between ns2 and ns3 from the tutorial of
>
> > ns3. I want to know which simulator will be useful for my research
> > area?
>
> I've used it for research on routing protocol security in OLSR and it's
> good. :) I've never used ns-2 except for some university assignments, so
> can't comment on that.

Please see also the University of Washington's work on wireless
jamming models for ns-3.
http://www.nsnam.org/wiki/index.php/NS-3_wireless_jamming_model

>
> Regarding ns-3's limitations itself for security research, one limitation
> I'm aware of is that we don't model processing delays within nodes, so you
> might have issues simulating something like a CPU DoS on a node.
>
> > 2. I have seen that ns-2.34 was released way back in 2009. Will there
> > be any new release of ns2 or the new development will always take
> > place in ns-3.

some ns-2 new releases (for ns-2.35) have been in work for a while now
but I haven't returned to these release candidates lately:
http://www.isi.edu/nsnam/dist/release/
although the cvs trunk of ns-2 is recently updated for gcc-4.6
compatibility.

not wanting to wait around for this release which has been brewing for
a while, someone started making Debian packages of the release
candidates:
http://packages.debian.org/unstable/main/ns2

In short, while ns-2 keeps up with some maintenance at a low level of
activity, and there will be an ns-2.35 release sometime soon, Lalith
is correct to point out that much more development and maintenance
work goes on with ns-3.

>
> 3. Are there any plan's to maintain the ns2 and it's documentation or
>
> > it is advisable to shift to ns3 for new projects? As the documentation
> > and books of ns2 are very thorough, I wish more tutorials, course work
> > and books should be published for ns3.

While I believe that ns-2 will continue to be maintained at least
lightly (to make sure that it still works as Tcl/Tk evolves and as
compilers evolve), there hasn't been much interest recently in
maintaining the models. When I've asked in the past, there haven't
been any volunteers to come forward to maintain its documentation (the
ns-2 manual). Despite this, I know that many in the TCP and wireless
sensor network communities still actively use ns-2.

I think it is advisable to use ns-3 for new projects unless there are
compelling reasons to use ns-2 (such as ns-2 having better model set
to start with for the research problem, or if one belongs to a
research group that has invested a lot in building around ns-2 over
many years and it still works well for the group).

>
> Both projects aren't exactly related (ns-3 is a complete re-write from
> scratch), but I can surely say that more development effort is going into
> ns-3 than ns-2, and ns-2 is only maintained rather loosely.
>
> > 4. Some protocols of ns2 are yet to be written for the ns3 (eg, DSR
> > and Dymo routing protocols). Are there any plans to develop a
> > Developer's Manual, similar to ns2 Manual for coding the new
> > protocols?
>
> What exactly would you like to see in such a manual? To develop routing
> protocols for ns-3, have a look at src/olsr, src/aodv or src/dsdv to see
> what you need to do. In essence, you'll need to create a class that inherits
> from Ipv4RoutingProtocol or Ipv6RoutingProtocol and extend it with the
> functionalities you require.
>
> For DSR, there's been an announcement here:http://mailman.isi.edu/pipermail/ns-developers/2010-July/008203.html
>
> > 5. Will there be any visualiser as NAM in ns2?
>
> There is PyViz, and NetAnim.

Some nam support for ns-3 will likely be added in the future, as well:
http://www.nsnam.org/contributed/ns-3-nam.tar.bz2

(if anyone wants to work on finishing this off, please contact me)

- Tom