larry@uunet.uu.net (Larry Lippman) (09/16/90)
In article <11739@accuvax.nwu.edu> den0@midway.uchicago.edu (funky chicken) writes: > In the process of recording information for billing, are there any > times when the originating or destination number is not recorded? It depends upon the type of central office, the type of message accounting apparatus (if not an ESS CO), and the policy of the operating telephone company for a particular CO or CO region. Nothing like a *specific* answer, huh? :-) Unfortunately, as far as I know, this situation really varies all over the map. > I suspect there are. For example, I recall that, in SxS offices with > ANI, each phone number's sleeve wire was cabled to a grid of bus > panels. Each directory number terminated on a card which held 10 > numbers. An identifier would come by and hierarchically (sp?) scan > this bus system for a 5800 hz which identified the calling number. Hey, you've been around! :-) Tone identification was only used in ANI-B; the later and far more common ANI-C and ANI-D all used 340 volt 200 microsecond pulses instead of the tone. > However, my memory tells me that all the directory numbers associated > with a PBX were fed into a single number network associated with the > primary or billing number. If this were the case, wouldn't ANI always > identify the calling number as being the primary number, regardless of > the actual line used? From what I have seen of ANI-C and ANI-D, there were two schools of thought that were *both* implemented: (1) send the actual line number and let revenue accounting sort it out; and (2) wire the ANI number network cards to send only a pilot billing number. When anyone ever asked my opinion, I always recommended (1) since it always provided more data in the event that troubleshooting was required. > When I used my modem > extensively, I had a service which allowed me to make unlimited calls > within my LATA for a a monthly fee of $25. Except for analyzing > traffic patterns, there would seem to be no reason to keep detailed > logs of calls in this sort of situation. Many operating telephone companies, including the RBOC's, record *all* of the call data in an ESS CO and keep it around for as long as six months on mag tape - although they might not publicly admit such a practice. This situation represents the ultimate in traffic usage recording and analysis. It's also pretty cheap insurance for billing dispute and toll fraud matters. This is really not that much data, and it's easy to sort out on a "straight line" basis for actual revenue accounting purposes. Consider that as a typical example, a 10,000 line CO making 20 calls per day per line (quite a bit of traffic, btw) will only generate between 10 and 15 megabytes of raw accounting data per day. That's for full, raw call data collection of even incompleted calls. One 6250 bpi mag tape reel can readily store 10 days or more data. > Hmmm. I suppose that it is, unfortunately, improper to refer to SxS > in the past tense. There's still a lot of SxS around in the boonies - even in RBOC territory! Larry Lippman @ Recognition Research Corp. {boulder||decvax||rutgers||watmath}!acsu.buffalo.edu!kitty!larry VOICE: 716/688-1231 || FAX: 716/741-9635 {utzoo||uunet}!/ \aerion!larry