Re: [ns] the time delay in node

[Lloyd Wood <l.wood@eim.surrey.ac.uk>]

On Mon, 29 Oct 2001, Ya Xu wrote:

> In (traditional wired)ns, queue is not the part of a node.
> The enQ and deQ trace happens on the LINK object, not
> a node object. (refer to ns doc)

very true.

but the question below is only asking about what is in the tracefile,
which ignores what happens inside ns. You wouldn't want want people
to gain the impression that ns can't simulate queuing...

L.

> I know many people are confused by this, but it is the
> way in ns.
> 
> In wireless simuation, a (broad) wireless node contains a
> queue inside it.
> 
> -Ya
> 
> On Mon, 29 Oct 2001, Lloyd Wood wrote:
> 
> > On Sun, 28 Oct 2001, Ya Xu wrote:
> > 
> > > On Wed, 24 Oct 2001, Sol Dashing wrote:
> > > 
> > > > We can assign a link delay in script file. However, I
> > > > am confused if there is a time delay in every node
> > > > simulated in ns-2. 
> > > > 
> > > > I check my trace file what I have got, and found that
> > > > the time of "+" and "-" corresponding to same packet
> > > > are always same. That means the time delay in each
> > > > node is 0, isn't it? In addition, I also noticed the
> > > 
> > > In the wired simulation, the answer is yes. 
> > 
> > If you have queuing, time delay in each node is not zero - even in
> > wired simulation.
> > 
> > http://www.isi.edu/nsnam/ns/doc/node211.html
> > 21.4 Trace file format.
> > 
> > L.
> > 
> > > You can find more details in the ns docs available in hte
> > > vint website.

<L.Wood@surrey.ac.uk>PGP<http://www.ee.surrey.ac.uk/Personal/L.Wood/>