You have no specific problem with your Tcl.
Since you have installed it, you only have to
provide the path to it during configure with the
following parameter:
./configure --with-tcl={your tcl8.4.5 path} --with-tcl-ver=8.4.5--with-tk={your
tk8.4.5 path} --with-tk-ver=8.4.5
What is meant with
I apologise, in my previous message I forgot to edit
the subject line.:-)
From the warning messages displayed, you have
installed ns-allinone 2.28, not ns-2.28 (the latter is when
you install it from the pieces). When you install it from the
pieces, it does not display such a warning.
The
You are not running the simulation from somewhere
where your environment variables for ns are not properly initialized
according to the NS-2. Init.tcl is one of the primary files executed during
the start of a simulation. Verify that you are executing the file from the
appropriate shell
The problem is that even though you have Cygwin
you do not have an X environment to execute X applications as NAM.
The solution to this problem, would be to install (in addition to cygwin),
the KDE ebvironment
for Windows, that co-operates with cygwin.
In this way you will have an X desktop,
The version I could use would be NS-2.28 at the earliest.
-Fk
-- Forwarded message --
From: raja priya [EMAIL PROTECTED]
To: NS ns-users@ISI.EDU
Date: Sat, 4 Feb 2006 22:49:42 -0800 (PST)
Subject: [ns] Ns2 version for Ad hoc networks
Hi
I am interested in working in
I would not agree with that, since for example AODV, one of the
most common routing protocols in MANETs uses UDP
to transmit its routing packets (RREQ, RERR, etc).
However, in order to avoid TCP/UDP you would have to
tamper with the specific application that accesses these
layers in the source
The point here is that you want to execute your local apps *through* your
nodes,
not independently *within* the program. In general, the command to execute
an
external program through an NS-2 script is to use the exec command followed
by
the program name and any parameters that it may need.
Yes, it is true that you may have installed it, but when you issue the
command
gcc --version then the system will report that you have the gcc version
4.0
installed. That is because in /usr/bin there is a symbolic link named gcc
that
points to the gcc-4.0 executable located in the same directory.
Well, the terms that you are referring to are
standard object oriented programming terms. There are some
sites that can offer a relative introduction to what
you are asking, but in order to understand in depth
these issues, you should read a good object-oriented
programming book, or an OTcl
Of-cource you can,
I have installed ns-2 in XP (with cygwin) and I can run it smoothly.
I also have installed both ns-2.29 and ns-2.28 in my Debian Workstation
and when I connect with ssh, (where no X servers are available), both of
them run smoothly.
No, you do not NEED an X server to run ns.
Yes. Check chapter 23 of the Ns-2 manual: Mathematical support
and specifically the part about Random variables.
Ns has a considerably good random number generator.
-Fk
-- Forwarded message --
From: [EMAIL PROTECTED]
To: ns-users@ISI.EDU
Date: Sat, 3 Jun 2006 13:56:38
As far as the random behaviour is concerned there is a quite good Random
number generator
implemented in NS, that can also generate random numbers from various
distributions.
You can check the NS-2 manual in the chapter Mathematical Support for
further information.
-Fk
On 7/13/06, [EMAIL
-
On 7/24/06, [EMAIL PROTECTED] [EMAIL PROTECTED] wrote:
-- Forwarded message --
From: Heo Heo [EMAIL PROTECTED]
To: ns-users@ISI.EDU
Date: Sat, 22 Jul 2006 17:18:26 -0700 (PDT)
Subject: [ns] Wireless packet size limit + wireless bandwidth question
Hi all,
I was
-
You might have already seen this site, but in any other case it might
be helpful.
http://www.item.ntnu.no/~wittner/ns/
It has ant-like code, but it is for ns-2.28. I think you might be able to
adapt it
to ns-2.29.
Hope that I have helped.
Regards,
-Fk
On 7/30/06, [EMAIL PROTECTED] [EMAIL
Dear all,
has anybody used, or tried to use the delaySamp_ object
that is provided with a queuemonitor object?
In the manual it states that this object can be used to measure queue delays
(page 76 of ns-manual).
It has two methods:
$queuemonitor get-delay-samples, which says it returns a
Dear ns-community,
I probably have found a way to answer my own question which was posted
previously concerning the calculation of average queue delay and size with
monitor objects.
After thorough study, I eventually found out what one has to do in order
to calculate the average queue delay and
-
It has several errors, which you can spot in the code that I send you, which
should work.
I also have included a simple trace output for you, so that you will be able
to see the packets that were actually
sent between the 2 agents.
As I understand from the code, you are trying to connect
Vidal [EMAIL PROTECTED]
To: Filippos Kolovos [EMAIL PROTECTED]
Date: Sat, 9 Sep 2006 14:24:20 -0300
Subject: Re: [ns] delaySamp_ object and Delay samples
Hi Filippos,
Did you solve your problem? Can you help me?
I'm trying to get Delay Samples from monitor-queue, with these commands:
set qm
-
In order to enable tcl debugging and debugging of the ns executable with an
external debugger (GDB, DDD, etc), you have to specify it differently during
the
configure stage of your ns-compilation.
They are different things.
So, in order to enable tcl-debuging you have to add to the configure
-
Please check http://mailman.isi.edu/pipermail/ns-users/2006-July/056470.html
for a description about this problem.
I also have attached a gzipped file with the patch applied to the code so
that it does not generate the error, called gt-itm-linux-patched.tar.gz.
Unzip, untart, re-make and
-
You can also check the link below in the ns-users archives
for calculating the average queue delay in wired links (i.e. packet delay)
using monitor objects.
http://mailman.isi.edu/pipermail/ns-users/2006-August/056736.html
Hope that I have helped.
Regards,
-Fk
On 12/30/06, [EMAIL
-
The -1 parameter means that the application will send an
arbitrarily large file (the file size will be selected randomly).
Check the NS-2 Manual, page 99, section 10.5.2 Starting the Agent
for more details.
Regards,
-Fk
On 1/5/07, [EMAIL PROTECTED] [EMAIL PROTECTED] wrote:
--
-
As far as I know, you just delete the ns-2 directories and
its uninstalled. After all ns-2 it's just an executable after the
./configure and ./make.
Regards,
-Fk
On 1/21/07, [EMAIL PROTECTED] [EMAIL PROTECTED] wrote:
-- Forwarded message --
From: Manoj Mistry [EMAIL
-
At first you need to schedule the loop proc to execute at the
start (or whenever you like) of the simulation.
Then, you need to put the re-scheduling code inside the procedure itself
in order to make it recursive. In this way the loop procedure will execute
once
at the start of the simulation
-
It does not have to do with the command that you give in the
tcl script, but with the generated command produced by the
$ns namtrace-all $filedesc tcl command, which gets written to a
file pointed by $filedesc.
The syntax of this command in the NAM file) is what has changed in later
versions
-
You could define the routing protocol to use
during the simulation, at the start of the tcl script
with the $ns rtproto command.
Then you could see what happens when changing the queue limits and
sizes.
NS, by default uses static routing with all routes pre-calculated
before the start of the
-
When you configure ns, you have to use the flag --enable-debug
in order to generate the appropriate symbols for debugging. Then
when you open the ns executable with DDD, you will be transferred
to the source files and you will be able to debug the whole source code
of NS-2.
Remeber though that
-
In order to debug TCL files you should use tcl-deug. In order
to debug the .cc files you need ddd (frontend for gdb).
Here is detailed web info on how to use both, from Pedro Vale Estela:
http://tagus.inesc-id.pt/~pestrela/ns2/ns2_debugging.html
Also, at the ns web site there is some info on
-
Hi there.
The solution to this problem is fairly simple, so do not worry.
At first, let me explain what causes this situation.
The problem does not lie at the Scheduler. The Scheduler works fine,
scheduling the event to occur at the appropriate time. The problem is
with the OTcl interpretor
-
I do not find any problem with your code.
I have executed it and as it seems to be working fine.
The problem might be that node 1 looks like it is sending
packets, but in reality it simply sends-out the previously enqueued
packets, at the time when the CBR app was functioning.
To verify my
-
Hi. I think that the problem that make reports has to do with the
fact that all the previously object files exist. You could erase the object
files of the .c files that you have modified so that make will have to
rebuild
only these files and not all of ns.
On the other hand, if this doesn't
-
Good morning
As it looks like from your code,
you are declaring the array n($i) only inside
your proc. However, because it is a local variable
it gets erased after the proc's execution. Therefore,
in your main script, when you try to access it with
$ns monitor-queue $n(3) $n(6) the
-
Hi,
I do not think that there is a problem with the variables per se,
but with the way that the TCL interpreter handles the command substitutions.
For example, when you invoke the command directly, i.e. with no
wrapping procedure, the NS scheduler schedules the event to occur
at time 1.0, but
,
Arvind
--
Filippos Kolovos
Software Systems Analyst Engineer
M.Sc. (Eng.) in Data Communications
Automation Networking Department
University of Macedonia Library
Egnatia 156, P.O.Box 1591
540 06 Thessaloniki, Greece
E-Mail: [EMAIL PROTECTED],
[EMAIL PROTECTED]
Profile: http
Ns-users@isi.edu
http://mailman.isi.edu/mailman/listinfo/ns-users
--
Filippos Kolovos
Software Systems Analyst Engineer
M.Sc. (Eng.) in Data Communications
Automation Networking Department
University of Macedonia Library
Egnatia 156, P.O.Box 1591
540 06 Thessaloniki, Greece
E-Mail: [EMAIL
35 matches
Mail list logo
