[Asterisk-Users] Softswitch
Brian D Heaton
bdheaton at c4i2.com
Fri Nov 7 22:18:59 MST 2003
On Fri, 2003-11-07 at 23:52, Eric Wieling wrote:
> As I understand it for large numbers of channels the TelCo usually
> provides a DS-3 or higher to the customer. The customer either connects
> that directly to their equipment (if their equipment supports it) or
> breaks the DS-3 out into multiple T-1 channels.
>
Often telcos will have DS3s between their facilities and then groom them
out to the switches with a 3/1 DACS. I much preferred having a DS3 from
each LEC CO into my 3/1 DACS over having a whole mess of T1s on physical
cross-connects. The good thing is that once the DS3 is up you don't
have to fool with transmission media testing for the next 27 T1s. It
also makes it much faster to turn up new interconnection trunks when
they can just do on channel-X of an existing facility rather than
engineering a whole new facility. Turning up new inter-switch trunking
was still an average of 90+ days though.
> As for the actual signaling it's either MF/DTMF (inband), PRI (out of
> band), or SS7 (out of band, usually only for TelCo-TelCo connections).
> Obviously your equipment has to support the signaling used. Most
> everything supports MF/DTMF and PRI, but not much supports SS7.
>
100%
> TelCos don't like to provide SS7 services. As I understand it, there
> isn't a whole lot of damage control facilities for SS7. So, if you have
> a piece of SS7 equipment that freaks out it can affect a large number of
> things all over the PSTN. A company with an SS7 connection to the
> another TelCo can do all sorts if interesting things like make calls
> that are billed to any number on the PSTN. IIRC, TelCos do require some
> kind of certification for SS7 equipment.
>
SS7 is kind of like the power grid. There are screening tables, but it
was built on the concept of everyone playing nicely in the sand box.
You *REALLY* don't want to know the nastiness that could be caused by
someone taking over a couple of core SS7 nodes on one of the larger
networks.
> I could be wrong about some of this, but I'm sure lots of people will
> correct me if I am.
More information about the asterisk-users
mailing list