Re: [ns] Satellite Problems

[Tarik Alj <aljtarik@cholla.inrs-telecom.uquebec.ca>]

sorry, typo.
------------- Begin Forwarded Message -------------

Date: Mon, 26 Jun 2000 11:40:32 -0400 (EDT)
From: Tarik Alj <aljtarik@cholla.inrs-telecom.uquebec.ca>
Subject: Re: [ns] Satellite Problems
To: L.Wood@eim.surrey.ac.uk
Cc: ns-users@isi.ed
MIME-Version: 1.0
Content-MD5: rnFbI85qaFQx5HRPK1EcIw==


>Date: Mon, 26 Jun 2000 15:41:42 +0100 (BST)
>From: Lloyd Wood <l.wood@eim.surrey.ac.uk>
>X-Sender: eep1lw@petra.ee.surrey.ac.uk
>To: Andrea Detti <a.detti@net.infocom.uniroma1.it>
>cc: ns-users mailing list <ns-users@isi.edu>
>Subject: Re: [ns] Satellite Problems
>X-url: http://www.ee.surrey.ac.uk/Personal/L.Wood/
>X-no-archive: yes
>MIME-Version: 1.0
>
>On Mon, 26 Jun 2000, Andrea Detti wrote:
>
>> I have to simulate a earth gateway for LEO satellites. Can the
>> sat-terminal object (used to simulate the gateway) link more than
>> one satellite?
>
>I'd have thought, after a glance at section 16.4 of nsdoc, that after
>you've done 
>
>set $satnode [$ns satnode-terminal lat long]
>
>you should be able to just do:
>
>$satnode add-gsl blah as usual.
>$satnode add-gsl blah for the second gsl link.
>
>...but I consistently get:
>
>phy.cc:82: failed assertion `channel_ == 0'
>
>even when adding an (unnecessary, if you see ns-sat.tcl)
>add-interface between the two add-gsl commands. At a guess, this seems
>to be due to an assumption about the relationship between channel and
>linkhead made in phy.cc. Anyone?

from the code in channel.cc, the channel maintains a list of interfaces (phy) to 
which it distributes packets (insertion is done using $channel addif $if). The 
assertion in phy.cc is there to make sure no channel is already present when 
initializing the interface (e.g. setting the targets), I guess. Could it be that 
there is some sort of redundance in the corresponding tcl code for sat? 

>
>Using a slightly modified 2.1b6-SNAPSHOT-20000220 here, with a current
>(i.e. hasn't been touched in six months) phy.cc.
>
>
>In any case, if that did work, note that both gsls will follow the
>same handover rules because there's only one linkhandoff manager per
>node rather than per link, which might get in your way. (Some
>per-link restructuring needed).
>
>A better approach might be to create two satellite nodes with the same
>latitude and longitude, and use them in parallel.
>
>
>> If it can, where I can find some example ?
>
>try adding an extra add-gsl line to the supplied iridium-ish and
>teledesic-ish scripts in ~ns/tcl/ex and see what happens.
>
>btw, users of the satellite code may find recent changes to
>http://www.isi.edu/nsnam/ns/CHANGES.html
>of interest.
>
>L.
>
><L.Wood@surrey.ac.uk>PGP<http://www.ee.surrey.ac.uk/Personal/L.Wood/>
>
>
>

Tarik 
------------- End Forwarded Message -------------