TELECOM Digest Wed, 14 Oct 92 23:55:00 CDT Volume 12 : Issue 780 Index To This Issue: Moderator: Patrick A. Townson Fibre Cables Not That Immortal (Don Kimberlin) Re: Fiber to the Home (David G. Lewis) How Should I Get My House Wired For Future Phone Needs? (Toby Nixon) Cost Accounting Software (Craig A. Brown) Different Rings For Different Things? (Justin Leavens) It's Back - FBI Wiretap Bill (Ron Dippold) ---------------------------------------------------------------------- Date: Wed, 14 Oct 1992 00:14:42 -0400 From: Don.Kimberlin@f730.n250.z1.FIDONET.ORG (Don Kimberlin) Subject: Fibre Cables Not That Immortal [originally posted in Fidonet BROADCAST conference, with some spelling corrections] Here's a just a bit of "real world" about fiber optic cable and those overblown statements about the apparent invincibility of fiber optic cable ... those things that trumpet phrases like, "Impervious to harm and interruption ... non-metallic and interference-proof!"... and so on. In fact, practical fiber optic cable of the type planted in the ground for intercity use has at least two important natural enemies, both of which have caused, and continue to cause failures of those cables, gophers and lightning. Both have been significant causes of failure of preceding cable technologies in the twisted copper pairs and coaxial cables of the earlier analog era. A recent edition of "Lightguide Digest" published by AT&T technologies tells quite openly how AT&T's RL series of fiber optic cable is constructed to attempt to avoid both kinds of natural enemies. First addressing the gopher threat, the article describes it as a "gnawing problem" that "is growing." It says, "the fact is well established that rodents such as gophers will gnaw through unprotected buried cables. Information collected in the 1950s indicated that gopehrs infested more than 60 percent of the continental U.S." An accompanying figure shows a distribution of gopher areas to include almost all the area west of the Mississippi River plys most of Florida, Georgia and about a third of Alabama. The article goes on, "To grind down incisors which grow approximately 12 inches per year, gophers spend 10 percent of their time, which means more than 90,000 bites per week, gnawing various items including fiber optic cable." The bite pressure exerted by a gopher on fiber optic cable is said to reach as high as 18,000 pounds per square inch. The habits of the gopher are described as a creature that keeps a home territory of several hundred square feet, saying that cables buried through gopher territories are the victims of repeated gnawing attacks. Reports have been made of gophers completely severing all-dielectric fiber cables in Ohio, an area not previously thought to be in the "gopher zone," and Colorado reports indicate gophers there are a widespread threat to unprotected fiber optic lines. Two methods are the only ones known to be effective against gopher attack: One is to place the nominal 1/2-inch diameter fiber cable in a 2-1/2 to 4 inch diameter rigid PVC conduit, an expense communications companies would rather avoid. The other is to use a cable with a stainless steel sheath around its fiber optic contents ... which makes the cable conductive, increasing the risk of lightning striking the cable and literally blowing the cable in pieces. Indeed, the article reports that one Mountain Bell fiber cut was caused by gnawing through a two-inch outer conduit, anyway. Larger cement or ceramic conduit might be used, but that would increase costs of placing cable evern more. To attempt gopher protection, AT&T has two calbe designs in its RL series. The first, for direct burial in rural areas, is called Primary Rodent- Lightning (Primary RL) Sheath. It consists of a corrugated stainless steel armor layer bonded to an outer polyethelyne jacket that contains a surrounding layer of wire strength members. The visual representation of this cable winds up looking very much like transoceanic deep-sea cables that have evolved over more than 130 years of submarine telegraphy around the world, the only difference being that the center of the cable is hollow, providing for a variety of forms of fibers inside. The other type, called the Lightguide Express Entry Rodent-Lightning LXE-RL Sheath, has but two strength- member wires in its jacket and a layer of "Water-Blocking/Lightning Tape" located under its stainless steel jacket. The article goes on to describe how such cable types are actually tested and rated for gopher resistance. A sample of cable is buried in a closed-off area (underground barricades to contain the gophers within a confined area) with ten gophers that have nothing else to gnaw on. Ratings of one to five indicate whether any given gopher managed to inflict damage ranging from cutting the outer jacket to completely biting the cable in two. The final rating of a given cable is the averaage of the scores of the ten different gophers placed in the area to attack the cable. The gophers used for the test are provided by the Denver Wildlife Research Center of the U.S. Department of Agriculture, and the gopher most often supplied is the common plains "pocket gopher" called Geomys Bursarius, a small animal only 7 ot 9 inches in length that rarely weighs more than half a pound. The other hazard that arises to intercity fiber optic cables, since metallic sheaths must be used to prevent gopher damage, and wires along the cable must be added to provide "strength" so the cable can be pulled and handled, is lightning. One of the lesser-understood factors about buried metallic conductors is that since they are often more conductive than the surrounding soil they are buried in, they are better, more attractive conductors to lightning than the dirt that surrounds them. This means that lightning can, and is, actually attracted to even buried cable, and that momentary lightning currents in the metallic members of those cable can rise high enough to actually melt a hole in the cable. Indeed, buried fiber optic cable can be struck in open rural areas, resulting in failure of the cable until the damaged section is replaces. One notable such case happened to MCI in a buried cable west of Jacksonville, Florida, where a lightning bolt that seemed to appear out of nowhere stabbed down into a buried fiber cable in an open field. Even more dramatic is the piece of damaged RL cable on display in the Charlotte, NC office of AT&T, which can be seen to have split open from the forces exerted on it when struck by lightning. That piece of damaged cable was struck with such force when it was buried four feet under an open soybean field near Davidson, NC. The cableman who repaired it stated that on the surface, a hole about the size of a golf ball had been made in the dirt, with spherical balls of dirt thrown back up around the hole where the lightning struck the ground on its way down to reach the metallic elements of the fiber cable. So, now you have some facts to bear in mind the next time you hear some salesman touting the apparently impervious, uniterruptible nature of fiber optic cable plant. It has the same sort of failure potentials that previous cables have had for decades! Don Kimberlin - via FidoNet node 1:250/98 INTERNET: Don.Kimberlin@f730.n250.z1.FIDONET.ORG ------------------------------ From: deej@cbnewsf.cb.att.com (david.g.lewis) Subject: Re: Fiber to the Home Organization: AT&T Date: Wed, 14 Oct 1992 17:40:46 GMT In article , Leonard.Erickson@f51. n105.z1.fidonet.org (Leonard Erickson) writes: > In TELECOM Digest V12#763 deej@cbnewsf.cb.att.com (david.g.lewis) > writes: >> Using your 5000 phone lines per (I'm guessing you mean) fiber pair, at >> 64kb/s for a phone line, you're talking sending 160Mb/s to each house. >> Let's use 155Mb/s, since that's an STS-3 SONET/SDH rate. Anyone know >> what an STS-3 FOT is going for these days? I don't have any >> up-to-date information (and if I did, I probably wouldn't be allowed >> to post it), but I'd guess that $30k/link (both ends) is correct >> within an order of magnitude. Even if using ring architectures and >> ADMs can drop your costs by a factor of two (unlikely to impossible), >> you're talking $15k electronics costs per house. >> Even if the cost of overlaying the fiber itself goes to zero, you've >> just incurred a $15k per subscriber incremental cost. At a 50,000 >> line CO, that's an investment of $750 million dollars. For capacity >> which will, basically, sit there until people figure out how to use >> it. > On the other hand, how *cheap* could a 64kb/s interface be? That's all > it'd take to supply a "regular" phone line. And in the quantities > involved, I'd expect the cost to drop *fast*. For the sake of > argument, call it $50. (I've seen RS-232 to fiber adapters in that > range). That's a different question than the one I was answering. I was responding to a statement to the effect that "capacity is free, so the telco should overbuild by a factor of 5000 because someday someone will find a use for it." My point was that capacity is most emphatically *not* free; even though the cost *for fiber* is no different whether one or 5000 equivalent phone lines is run over that fiber, the cost of the fiber is by no means the only cost factor -- and may not even be the dominant cost factor -- in determining the cost of providing large amounts of capacity to a user. We could probably discuss for hours (and people do, regularly) the economics of fiber to the home/fiber to the curb/fiber in the loop. I'll just make some observations: * Most RS-232 to fiber converters use multimode fibers and LED transmitters, which are considerably less expensive than semiconductor lasers for use with single-mode fiber. And if you're building to allow megabandwidth in the future, you'll build single-mode fiber. * Depending on whose studies you believe, fiber in the subscriber loop as far as a pedestal serving four to eight residence customers, with copper drops from there to the demark, is somewhere between 25% above (and dropping) to 25% below total copper distribution, on an installed first cost (IFC) basis. This architecture has the advantage of pushing fiber into the loop to a certain point, with the possibility of running it the rest of the way to the residence when the economics improve. * Also depending on whose studies you believe, fiber to the home is anywhere from three to ten years away from matching copper on an installed first cost basis. * Regardless of whose studies you believe, all these IFC comparisons omit from consideration cost and other factors both for and against fiber, like cost of operations support system upgrading (against), improved operational capabilities of fiber (for), improved quality and reliability (for), ease of upgrading (for), and the fact that the copper distribution plant is already there (big against). * Telcos will generally act on whoever's studies *they* believe. Ameritech has placed a large order for fiber in the loop equipment, which leads me to believe that they believe that it's cheaper than copper in some scenarios. In general, you can expect that telcos are somewhat intelligent when it comes to building networks, and if something's cheaper than what they're doing now, it won't take them *too* long to latch onto it. > And if you need extra capacity, instead of running more lines, you > replace the interface box at the user premises. Indeed, this is one large advantage of fiber. > I'm not sure whether the CO end of the fiber would be better served > by a box capable of handling a range of line capcities, or by a > dedicated box. A "box" capable of handling a range of capacities without changing any hardware is, currently, a myth. Some systems (like AT&T's DDM-2000) use one system framework (processor, low-speed interfaces, common elements, etc.) with several different varieties of optical interface for different line speeds (e.g. OC-3 and OC-12), but if you change the line speed, you have to at least change out the optical interface. > Either way, at some point, in the course of upgrading, it'll be > necessary to "move" the fiber to a different box. While this is not as > simple as splicing copper, I doubt that it'd be *that* expensive. Actually, done right, it's easier than splicing copper; if done right, the outside plant (OSP) fiber terminates in a splice tray and is spliced onto a connectorized jumper; the jumper is then connected to whatever equipment you want. You want to change the equipment, you move the jumper. Of course, you have to coordinate things so you don't bring the user down ... Typical Disclaimer: I don't work with this stuff, and I never did at AT&T, so I don't necessarily know what I'm talking about, and even if I do, it's not necessarily what my company's position is. Anywhere. Ever. ------------------------------ From: Toby Nixon Subject: How Should I Get My House Wired For Future Phone Needs? Date: 15 Oct 92 01:30:06 EDT Organization: Hayes Microcomputer Products, Norcross, GA My wife and I are considering building a new house (well, having it built for us) for ourselves and our five kids. Before we know it, those kids are going to be teens, with heavy telecom demands. Although we currently have just two standard POTS circuits, I want to look ahead and get the house wired up-front to handle our needs far into the future (maybe a mini-PBX, maybe ISDN, an intercom, whatever). What I'm think of is having three or four-pair twisted pair cable run from each room separately (not daisy-chained the way they normally do it) to a punch-down block in some central location, so we can hook things together anyway we want. Is that a good idea? Do you have any other advice for us? Thanks in advance. Toby Nixon, Principal Engineer | Voice +1-404-840-9200 Telex 401243420 Hayes Microcomputer Products, Inc. | Fax +1-404-447-0178 CIS 70271,404 P.O. Box 105203 | BBS +1-404-446-6336 AT&T !tnixon Atlanta, Georgia 30348 | UUCP uunet!hayes!tnixon Fido 1:114/15 USA | Internet tnixon@hayes.com ------------------------------ From: craigb@craigb.uark.edu (Craig A. Brown) Subject: Cost Accounting Software Organization: University of Arkansas Date: Wed, 14 Oct 1992 15:47:38 GMT The University of Arkansas is presently using the TeleMate call accounting software from Complementary Solutions for capture and billing of institution long distance calls. We are considering providing our own, billable service to students. Based upon the length of time it currently takes to generate billing and management reports from TeleMate on a 486/25, we are somewhat concerned about the impact of the additional load. We have received material from Styker System Incorporated for their MegaBase product which they claim is the world's fastest large database cost accounting software. Have any of you had any experience with this product? Thanks in advance for any information. ------------------------------ From: leavens@mizar.usc.edu (Justin Leavens) Subject: Different Rings For Different Things? Date: 14 Oct 1992 10:25:54 -0700 Organization: University of Southern California, Los Angeles, CA CLASS features such as the "Repeat Dialing" and "Priority Ringing" that are being offered now. I am curious to know if there are any specifications or standard in place so that these special rings could be picked up by a modem (one in the future specifically designed to look for these rings) or a terminal communications program. Justin Leavens Microcomputer Specialist University of Southern California ------------------------------ From: rdippold@qualcomm.com (Ron Dippold) Subject: It's Back - FBI Wiretap Bill Organization: Qualcomm, Inc., San Diego, CA Date: Thu, 15 Oct 1992 00:28:53 GMT According to {EE Times} this week (Oct 12.), the FBI has sent Congress another version of the legislation to make it easier for them to listen to any phone conversation. "Both the House and enate are expected to act quickly on the bill when Congress returns early next year. "The bill is a slightly rewritten version of legislation submitted last spring, which was dropped after both lawmakers and the telecommunications industry complained." Then there's more whining from the FBI that they're apparently incompetent to handle wiretapping with digital technology, Unfortunately, the "slightly rewritten" isn't detailed, so there's no specifics as to what changed. It goes on to say that industry is opposed, mostly because of the cost of fitting and retrofitting everything so that the FBI can spy on people. The FBI claims it will "only" cost "$300 million, or around 1.5 percent of the industry's total yearly acquisition budget." Any government estimate of taxation is going to be way too low, especially when applied to an industry with the resources of the telecom industry, so we apparently have a lower cost floor of $300 million. "Industry has been talking to the FBI to try to get some agreement on a private level rather than through legislation. But the FBI says these talks are moving too slowly and it wants its legislation to go ahead." I can't see any of the three major candidates vetoing something like this once they take office, so we may have to swamp Congress again. ------------------------------ End of TELECOM Digest V12 #780 ******************************