TELECOM Digest Tue, 7 Feb 95 01:29:00 CST Volume 15 : Issue 82 Inside This Issue: Editor: Patrick A. Townson Re: LD Termination Fees to RBOCs (Fred R. Goldstein) Re: Cellular Fraud: How Much of it is Real Money? (Michael D. Sullivan) Re: What is a T1 Line? (Al Varney) Re: Cheap Way to Get an 800 Number? (Paul Robinson) Re: 28.8k bps Modem (Paul Robinson) New Archives Email Service Feature (TELECOM Digest Editor) TELECOM Digest is an electronic journal devoted mostly but not exclusively to telecommunications topics. It is circulated anywhere there is email, in addition to various telecom forums on a variety of public service systems and networks including Compuserve and America On Line. It is also gatewayed to Usenet where it appears as the moderated newsgroup 'comp.dcom.telecom'. Subscriptions are available to qualified organizations and individual readers. Write and tell us how you qualify: * telecom-request@eecs.nwu.edu * The Digest is edited, published and compilation-copyrighted by Patrick Townson of Skokie, Illinois USA. You can reach us by postal mail, fax or phone at: 9457-D Niles Center Road Skokie, IL USA 60076 Phone: 500-677-1616 Fax: 708-329-0572 ** Article submission address only: telecom@eecs.nwu.edu ** Our archives are located at lcs.mit.edu and are available by using anonymous ftp. The archives can also be accessed using our email information service. For a copy of a helpful file explaining how to use the information service, just ask. ********************************************************************** *** * TELECOM Digest is partially funded by a grant from the * * International Telecommunication Union (ITU) in Geneva, Switzerland * * under the aegis of its Telecom Information Exchange Services (TIES) * * project. Views expressed herein should not be construed as represent-* * ing views of the ITU. * ********************************************************************** *** Additionally, the Digest is funded by gifts from generous readers such as yourself who provide funding in amounts deemed appropriate. Your help is important and appreciated. A suggested donation of twenty dollars per year per reader is considered appropriate. See our address above. All opinions expressed herein are deemed to be those of the author. Any organizations listed are for identification purposes only and messages should not be considered any official expression by the organization. ---------------------------------------------------------------------- From: fgoldstein@bbn.com (Fred R. Goldstein) Subject: Re: LD Termination Fees to RBOCs Date: Mon, 6 Feb 1995 14:47:44 GMT Organization: Bolt Beranek & Newman Inc. In article rta <75462.3552@compuserve. com> writes: > Several issues are being raised here so let me try to sort them out. > For FGD, many tariffs if you break them down, have a 1/3 the cost in > switching, paying for COs and access tandems, 1/3 in transport, the > bandwidth, and 1/3 in Common Carrier Line on the originating or > terminating end, what the carrier pays to support (subsidize) the line > to your home or business. This breakdown is approximate and varies > from LEC to LEC, but it gives you the overall picture. Only about a > third of the charge is for bandwidth as most of us would interpret the > word. The other element is switching. That happens to be the way Feature Group trunks are priced. Of course the whole trick to FG pricing is that it's intended to be "contributory" (profitable). This comes from the old separations game of "splifs", for "subscriber plant factor" (SPF). In that game, the average usage of LOCAL lines is divided into interstate and intrastate baskets. The interstate portion is MULTIPLIED by SPF and then the total cost is divided amongst the two jurisictions. Given SPF of 3 (old ballpark; I don't know what it is now), then if 15% of calls in a jurisdiction were interstate, then the cost would be divided 85:45 to local/interstate. The subsidy, folks, is in the splifs. Interstate cost is currently divided into the part paid via tolls and the part paid via CALC ($3-6/mo "access charges"). The toll-usage part is divided into different components and adds up to 3-5c/minute/ side-of-call for most telcos. > Switching and the lines to the customer premises are really what is > expensive. The lines are not heavily utilized, in the case of many > small businesses and residences. Switches are expensive and are not > set up to support multiple hour calls. Regular business calls that > get through are typically 4 to 6 minutes in length with residential > calls longer since many are placed to friends and family. Switches > were engineered on the assumption that most calls would be short. As > data usage increases, the switches have to get bigger, an expensive > proposition or the multiple hour data calls will have to shifted to > another technology, such as packet or cell switching where switches > and long distance circuits are not tied up during think time. Strowger switch costs were heavily usage-oriented. Modern switch costs have a minor usage component, less than .1c/minute if you compute it, for local calls. Data usage could "pay its own way" at a fraction of a cent/minute for local calls. Don't let the phone companies fool you otherwise. > Most modern CO switches detect a phone that is off the hook and not > transmitting and generate an obnoxious tone to get you to hang up. Proof that somebody's trying to confuse the issue. The howler tone does not tie up major resources in a modern switch; the line card is just connected to a tone generator channel. On an old Strowger it could hang a line finder. > Jerry Harder Senior Partner > Renaissance Telecommunications Associates Lemme guess. A consulting firm specializing in helping telcos win rate hikes? Fred R. Goldstein k1io fgoldstein@bbn.com Bolt Beranek & Newman Inc., Cambridge MA USA +1 617 873 3850 ------------------------------ From: mds@access.digex.net (Michael D. Sullivan) Subject: Re: Cellular Fraud: How Much of it is Real Money? Date: 6 Feb 1995 03:50:31 -0500 Organization: Wilkinson, Barker, Knauer & Quinn (Washington, DC, USA) bruce@zuhause.MN.ORG (Bruce Albrecht) writes: > My question is that if they are truly losing hundreds of millions of > dollars a year to fraud, why aren't they switching to known technologies > (e.g., GSM based) which have per call authentication using a random > number query with an encrypted key response, when such systems have > been available for several years in other parts of the world? If they > really are losing $300 million or more a year, it must be cheaper to > replace every single cellular phone with a more secure system than to > let these losses continue to escalate. At least in part because the FCC *requires* all cellular carriers to provide AMPS-standard, unencrypted, analog service. They are free to offer alternatives as well, such as CDMA and TDMA (GSM, too, if anyone cared to do so), but they still have to provide service to unencrypted analog customers. Also, phones using the unencrypted AMPS standard are cheaper and provide better voice quality than the alternatives; the companies' customers have analog phones. Cutting off the majority of your customers to prevent fraud is a great way to go out of business. The cellular manufacturers have been working on new standards for nearly ten years. They aren't about to come up with yet another standard -- analog with encryption -- that will be incompatible with every system out there, and the carriers aren't about to buy it when they plan to transition to digital over time and do away with analog when the FCC allows it and their customers accept it. PCS, on the other hand, has the advantage of reinventing the wheel, since there's no embedded base of equipment. Unfortunately, there are something like seven different standards under consideration for PCS (including at least one GSM variant). > If most of this amount is funny money, "lost profits" that they never > really expected to generate, and use of excess capacity, then are the > phone companies crying wolf? Are we currently in the position where > the phone companies are like the suburban/rural household that never > locks their doors "because crime never used to be a problem", and now > screams for more police because they keep getting burglarized, but > still never lock their doors? The carriers have tried a lot of things, from "electronic signatures" of phones that have to match a database entry to PIN numbers etc. The latest is the FCC's new rule, which the manufacturers hate, that requires the ESN to be unchangeable, period. This, unfortunately, is kind of like making it illegal to build a house without locks to prevent burglary. > My main concern over cellular telephone fraud is that because it is > partly due to decisions made by the phone companies, and that it's > probably been exaggerated, that our government is either going to > respond with excessive legislative and/or regulatory reaction to a > technical problem, or with no action at all. Either way, it sends the > wrong message. As discussed above, it's not just decisions made by the phone companies; it's decisions made by the government, purchases made by consumers, and the limits of technology. Are we going to eliminate credit card fraud by eliminating account numbers or mag stripes and require voiceprints? No. Are we going to eliminate software piracy by reqiring copy protection? No. Similarly, we can't eliminate cellular fraud by requiring use of a standard that obsoletes all of the existing phones and cellular systems. Michael D. Sullivan | INTERNET E-MAIL TO: mds@access.digex.net Bethesda, Md., USA | also avogadro@well.com, 74160.1134@compuserve.com ------------------------------ From: varney@usgp4.ih.att.com (Al Varney) Subject: Re: What is a T1 Line? Organization: AT&T Network Systems Date: Mon, 6 Feb 1995 08:54:21 GMT In article , Butch lcroan/.nameBalcroan Lilli wrote: > ARGHHH !! I am really getting tired of this BIT-Robbing conversation > that Jeffery Rhodes started. I used to work with Jeffery and he > certainly is a smart guy, but he is no expert in this area. > ... but there is more to consider here. The " MU LAW " is not a linear > scale it is more of a log function with more steps closer to the lower > levels where the ear is more senstive. I really can't believe 2 DB; > come on Jeff, 3 db is half power *and also the least amount the ear can > detect*. Jeff certainly can fend for himself in this forum, so this is just to add a little FACT to the fray. Butch, you were right -- Jeff was stretching the truth. The actual drop in Signal-to-Distortion ratio with 1/6 bit robbing between two switches is 1.8 dB, not 2. Of course, when there are 5 switches in the connection, the ratio drops by up to 4.2 dB. (5 out of 6 bits are affected on 10% of the calls, 4/6 on 40%, 3/6 on 40%, 2/6 on 10% and 1/6 less than 1%.) On a typical mid-length call (intraLATA toll or between adjacent LATAs), 3 out of 6 bits will be robbed 57% of the time and the ratio drops by about 3.2 dB. > I really doubt that the modems are affected by this as > much as Jeffery has stated. I would more believe than something more > common such as " ECHO " and several other more common impairments are > really alot more important than a occasional bit robbing. I agree that S/D ratio is not a BIG issue with modems, but it does place a small per-call variation into the connection -- and that could be just enough (combined with echo/cross-talk) to force a modem to back-off to a lower transmission rate. It is probably at least as much of a factor as the individual variation between two different modem lines at a common location. > There are also several new technologies such as fiber that have > introduced timing impairments such as " Jitter " into the equation. Jitter exists on all synchronous transport mechanisms, even ISDN lines. Copper-based, fiber-based, microwave-based, soliton-based all jitter. Whether it is a significant problem usually depends at least partly on the bit-rate, repeater design and bit-detection mechanisms. You seem to have it in for optical-based media -- what do you have against photons? Al Varney ------------------------------ Date: Mon, 06 Feb 1995 10:11:14 EST From: Paul Robinson Organization: Tansin A. Darcos & Company, Silver Spring, MD USA Subject: Re: Cheap Way to Get an 800 Number? > Some friends and I are starting a new small business. We would like to > have an 800 number. How do I get one? You call a long distance carrier's 800 number and tell them you want one. Within a couple of business days they will turn on the number. > Other than ATT/MCI/Sprint, are there other people who can provide an > 800 number cheaply? Don't know how much you mean by 'cheaply'. My 800 number from AT&T costs me $8 a month plus usage, which is typically around 20-25c a minute depending on how far the person is from Maryland. As I typically have a low calling volume, the costs usually run only $12 a month or so. > How do I minimize my cost? First you have to figure how many calls you expect to receive. If you are doing a substantial number of calls now, you might see some of them move to the 800 number. If you see you are getting large volumes of calls, you can change your service to a different plan which charges slightly more for the service and less per minute. > How do I get 800-CALL-MY-BUSINESS? Do I have have to pay extra for a > "good" 800 number. I originally had Sprint for my 800 number. The number I wanted -- since my company name is "Tansin A Darcos & Company" -- the number I wanted to get was 1-800-TDARCOS. I couldn't get it; it was apparently reserved by someone else, even though calling it indicated that the number said it wasn't in service. The business line here is 301-587-6354. So, I got *that* number from Sprint, e.g. 1-800-587-6354 which is a nice idea. I checked later with AT&T when I saw their rate was fairly competitive with Sprint. Guess what: AT&T *was* able to give me 1-800-832-7267 (800-TDARCOS). So I moved to them. There was one item that I was not told, which suggest you ask all costs in advance. Sprint enabled the number I asked for -- the one that matched my telephone number -- for the fee of something like $10 a month, and included a listing with 800-555-1212 and even allowed calls from Canada; they asked me if I wanted to allow that, I said ok. When AT&T turned on the new number, there was an additional "installation" charge of $45. During a one month period, both 800 numbers from each carrier terminated on the same 301-587-6354 number so, unless a carrier has some rules against it, you can have multiple numbers terminating on the same line. Also, there is no longer an instate/out of state restriction. I can call my 800 number from another phone in the same room, or I can call it from Virginia (which is out of state but also a local call in this area). You no longer have to have separate instate/out of state numbers. And my bill shows the ANI of every call I've received, which occasionally includes calls from places like Texas or New York. Another thing to consider is who had your 800 number before; if it was in use by someone else recently. I used to have a big problem with my old 800 number I had a couple of years ago (this was before you could ask for a specific number); the former owner was a freight broker, so I got calls from long hall truckers across the country telling me about partially empty trucks available from point a to point b. I finally had to put an announcement on the line that said "Tansin A. Darcos & Company, a computer software development company." This is a point that is probably going to be a problem for the local telephone companies. Bell Atlantic will sell an 800 number for instate calls (it mentions this in the phone book); because of the long distance restriction they cannot provide interstate delivery. I doubt they can be much cheaper than a national carrier; if I can get an 800 number from Sprint for $10 a month that allows me to receive calls from anywhere in the US or Canada, why would I want a number that only works in my home state? About the only reason I could see is if you wanted a restriction on incoming calls from this state. But that's a problem too: within a 50 mile radius of where I am includes in MD: Baltimore, BWI Airport, all of Montgomery and Prince George's county, the state capital at Anapolis, parts of Howard County, but it also includes interstate areas including all of Washington, DC, Independent Cities of Alexandria, Falls Church, and Fairfax, Va; towns of Vienna and Manassas, plus Fairfax and Arlington Counties in VA, plus Washington National and Dulles Airports. If I want to reach anyone I can get to in a two-hour radius, an in-state only 800 number is worthless. In fact, most companies operating anywhere within twenty miles of Washington typically will operate in DC plus the cities and counties in both states surrounding it. For example, a large local plumber here lists one number to call in the greater Washington area: 1-800-4-HOT-WATER. I know they are licensed in all three jurisdictions. While they could probably take calls from Baltimore, or Arlington, VA, calls to them from, say, Dallas/Ft Worth or Los Angeles or Chicago would be worthless. Yet since they don't advertise there, paying more per month to block calls outside the local area probably isn't worth the extra cost, e.g if you get $3 worth of wrong numbers a month, paying an extra $20 to restrict area codes you don't want to service is not cost effective. ------------------------------ Date: Mon, 06 Feb 1995 10:54:58 EST From: Paul Robinson Organization: Tansin A. Darcos & Company, Silver Spring, MD USA Subject: Re: 28.8k bps Modem Victor Hu , writes: > I just purchased a 28.8 K modem with the brand "Supra". I paid extra > to get the 28.8 K instead of the 14.4 K. Probably about twice as much. 14K modems are down to $50 or so. 28Ks are probably in the $100 range. I purchased a 14.4 two years ago and it cost me $215. Consider yourself fortunate. Mine still works quite úÿ well -- I am using it to enter this message -- and like most modems, will be obsolete long before it wears out. If my Internet provider decides to upgrade to 28.8K, I will probably get one. Otherwise, since I can't use one yet, I'm not going to bother. > 1. Is the bps across the twisted pair wire actually running at 28.8 or > 14.4 when 28.8 is invoked? Or is it just data compression? The raw data rate for a modem will be from 110 to 28,800 baud (or 14,400 baud) depending on what the other side agrees on. The rate will be the lowest of whatever the two modems agree on. If you call up a service that has only 14.4 modems, or 9600 baud modems, or even 2400, you will only get 14.4 or 9600 or 2400 even though your modem can do more. If both modems are 28.8 and both have their highest speed enabled, you should see 28,800 baud before any compression occurs. The data is not sent at 28,800 bits per second, however. Typically the modem will divide up the telephone line into six or more channels, and run each channel at 2400 to 4800 bits per second. By multiplexing six channels at 2400 baud, you get 14,400 baud, etc. > 2. What kinds of host supports 28.8K? I only connect up to my > university's computer which only runs at 9.6K max. Your university may have so much load they can't run faster than 9600, or the terminal controller might not be able to handle it. Many mainframes can't do I/O faster than some otherwise slow speed by comparison. When a IBM Sierra mainframe came out, a 9600 baud modem probably cost as much as an ISDN BRI interface does now; hundreds or thousands of dollars, so the port controller was probably set up with that as the maximum. The school might not have money in the budget to upgrade modems, or the hardware might not be able to support those kind of speeds, or it could be the administration was waiting until the 28,800 speed was standardized by ITU, as some modems used proprietary methods to communicate about 14.4 and thus you might need the same brand at both ends. You might ask them if they plan to upgrade to 28.8 now that the ITU has standardized the method of delivering 28,800 baud. > 3. What is the speed of fax machines? There are two speeds for transmissions. First, when the connection is being set up, each side will send an identifier sequence. I call it the "answerback" after the similar sequence sent by a telex machine. This identifier sequence is called a TTI or CSI. One of these will typically appear in the log that the fax machine prints after 20-40 transmissions indicating the identifying machine. The other is the telephone number or other identifier that appears in the display window. The two items may be different. This information is transmitted by each machine at 300 baud, which is okay since it is typically no more than 60 characters for each side. The sending machine then increases its speed and the transmission takes place in the equivalent of "half duplex" mode, except that the recipient machine typically acknowledges the end of each page and end of transmission. The ITU standard for fax machine transmissions supports 4800, 9600, 12000, and 14400 baud, but typically a fax machine that does printing will do 9600 tops, and can be downgraded to 4800 if line conditions are bad. 12000 and 14400 are typically for fax modems in computers. > 1. The Supra has a nice display (external version for the PC) that > shows the mode of transmission. > 2. However, I found that it required a different initialization string > than that suggested as default for modems that are Hayes compatible. The strings for each modem are because they all do different things, and thus, to enable those features you have to set certain values. For example, you can do a feature called "port locking". Currently, my terminal program sends and receives data to and from my modem at a "locked" speed of 19200 baud. The modem will transfer data to and from the computer at 19200 baud, whether the connection at the other end is 110 all the way to 14400. The typical rule is to indicate every connection is at 19200. I can set a switch register and the modem will also report the actual connection speed. I can set a switch and have it enable or disable data compression, and I can enable or disable error correction. And I can set switches so that it tells me whether the other side allows or does not allow compression or error correction, or force error correction or compression. For example, when playing DOOM over a modem, the modem will run at 9600 baud (later revisions of the driver support 14,400). The modems must not use data compression or error correction (because the extra time to do this can lose synchronization between the two computers), and the setup string for my modem will specifically disable these features. At the end of the game, the driver will issue an AT Z HO to disconnect the line and reset the modem to the default settings. If I have caller ID service on a line, I can enable the modem to send the data after the "RING" message. I can enable the data in hexadecimal display digits, or I can enable it as ASCII text. I can also tell the modem to resend the last Caller-ID string it got, and I can tell it to resend it in hex or ASCII, even if I had already received it using the other mode. I can also turn off reception of Caller ID data. If I want the modem to send or receive facsimile data, the program must use the AT&F prefix with certain commands to tell the modem to either place a fax call or receive a fax. Every feature in a modem requires controls on it in order to enable or disable them as needed in a particular instance or application. Oh, yes, one more thing. The alleged claims of data compression giving throughput rates of 50+K and 110K on 14.4 and 28.8 modems is sheer fantasy. If you were sending a 100,000 byte file consisting of all spaces, or all the same character, you might see those kind of rates. On my 14.4K modem, on ASCII text files, if I use a locked port at 38,800 baud, and enable compression I can see transfer rates of as much as 3000 cps in rare cases, and typically around 1800 cps if the data compresses well. For binary and compressed ZIP archive files, I have generally seen average transfer rates in the 1600-1620 cps. For a 28.8K modem, you can expect to probably see rates around 3200-3400 cps, depending on the content of the material, if the other side uses a 28.8K and can stuff it fast enough to keep the line loaded. MCI Mail supports 14,400 baud on their dialins, but I typically see rates in the 700-1000 cps rate, probably because their VAX machines are heavily loaded. Occasionally I'll see rates as high as 1200 cps. ------------------------------ From: TELECOM Digest Editor Subject: New Archives Email Feature Date: Tue, 07 Feb 1995 01:00:00 CST Monday evening I installed a new feature in the Telecom Archives Email Information Service software called AREACODE. This allows the user to enter one or more areacodes and get back email telling where the code is located. This is just an additional command installed in the software, used in the same way as the other commands. You can insert as many areacodes as desired and get back responses on each one. If you are not already familiar with the Telecom Archives Email Information Service, you can request a help file explaining how to use the service. Send email to tel-archives@lcs.mit.edu. The subject does not matter. In the text of your message, at the left margin, enter these commands on your first use of the program: REPLY yourname@site HELP INFO END The REPLY command *must* be first, and END *must* be last. What you put in the middle depends on what you are seeking. The help file you will get in return explains the other commands. To use the new AREACODE command, it would look like this: REPLY yourname@site AREACODE 208 701 302 509 (or whatever areacodes you want to check). END Watch for a response in email, typically a minute or two later. Other interactive commands you can use are: SEARCH to search the back issues of the Digest for subject and author names since 1989 forward. GLOSSARY to search the several glossaries on line in the Archives checking for abbreviations and telecom terms you want to learn about. The help file will explain it all when you order it and INFO. Carl Moore keeps the area codes files up to date, so blame him, not me if you don't find all the newer codes there. Patrick Townson ------------------------------ End of TELECOM Digest V15 #82 *****************************