Joe Greco wrote:
Stephen David wrote:
i don't have a specific bug in mind, i was just wondering WHY call progress doesn't work so well -- in particular, on analog lines. ie. is it a hardware or software problem (or both). with more info, i'd like to help to work out the kinks, for myself and everyone. :)
Back in the days of Stowger exchanges you knew when the called party
answered, by a reversal of the DC voltage on your analogue line. With
digital exchanges that stopped, and no solid feedback is given to the
caller on ordinary analogue lines. You have to infer that someone has
answered, and the reliability of that can be poor. Digital lines, like
ISDNand SS7, and protocols like MFC/R2 tell you positively that someone
has answered.
That's a good explanation. I'll expand upon it a bit by saying that even
with reversal, there's a limited amount of information you can represent
with that. POTS was always intended to be cheap basic phone service, and
keeping it simple was not considered a downside by the phone company.
As it is, you run into an information representation issue with the
existing technology: the entire "traditionally used" bandwidth of the
channel during a call is used for audio data (that is, to say, that they
send an analog signal). As a call originator, you really can not tell
the difference between a ringing signal generated by the phone company
and a ringing signal caused by the called party picking up the phone and
playing an identical sound. Reversal fixed that, but was largely made
obsolete by out of band supervision - since the real purpose of reversal
was for the telephone company to be able to bill correctly for completed
calls (IIRC, ICBW).
Actually it was not really intentional. The reversal back to the calling
party was just a byproduct of the way a Strowger exchange worked. Within
the network it was used for billing purposes.
More difficult is the problem of knowing when the remote end has gone
away. Reversal, loop break, dial tone, and just plain silence are not
all that unusual as methods of detection. In some cases, you do actually
need to infer that the remote has gone away.
Hangup is relatively easy. Most lines now give a strong distinct beeping
either the moment the phone is dropped, or a short time after. The
problem in * is its detector is not very good, or very voice immune. I
have a much better one in my spandsp library, but it isn't integrated
with * right now. Detecting answer is the tough one. There is nothing
unambiguous about it.
There's no real excuse for us to be using this technology anymore, with
the availability of things like ISDN BRI, which allows for digital
signalling of call progress. However, we continue to use it because the
ILEC's have done such a fab job of making ISDN a dead technology. Funny
thing is, it'll end up biting them where it hurts, as customers drift to
VoIP to gain the features that ISDN promised, at a fraction of the cost.
As someone whose colleagues built one of the first ISDN muxes in the
80's, I can tell you attitudes made it dead from day one.
(I say that as someone who currently brings in dialtone on BRI, btw)
Steve
_______________________________________________
Asterisk-Users mailing list
[EMAIL PROTECTED]
http://lists.digium.com/mailman/listinfo/asterisk-users
To UNSUBSCRIBE or update options visit:
http://lists.digium.com/mailman/listinfo/asterisk-users
