A toll free call paid for by the called party, rather than the calling party. A generic and common term for In-WATS (Wide Area Telecommunications Service) service provided by a phone company, whether a LEC (Local Exchange Carrier) or an IXC (IntereXchange Carrier). In North America and in order of their introduction, all these In-WATS services have 800 (1967), 888 (1996), 877 (1998), 866 (2000), or 855 (2001) as their “area code.” (Note: Future 800 numbers will follow the convention 8NN, where NN are specific numbers which are identical. Such 800 service is typically used by merchants offering to sell something such as hotel reservations, clothes, or rental cars. The idea of the free service is to entice customers to call the number, with the theory being that if the call was a toll call and therefore cost the customer something, he or she might be less inclined to call. Supplies or 800 services use various ways to configure and bill their 800 services.

800 Service works like this: You’re somewhere in North America. You dial 1-800, 1-888, 1-877, 1-866 or 1-855 and seven digits. The LEC (Local Exchange Carrier, i.e., the local phone company) central office sees the “1” and recognizes the call as long distance. It also recognizes the 8NN area code and queries a centralized database before processing the call further, with the query generally taking place over a SS7 (Signaling System 7) link. The centralized database resides on a Service Management System (SMS), which is a centralized computing platform. The database identifies the LEC or IXC (IntereXchange Carrier) providing the 8NN number. Based on that information, and assuming that the toll-free number is associated with an IXC, the LEC switch routes the call to the proper IXC. Once the IXC has been handed the call, it processes the 800 number, perhaps translating it into a “real” telephone number in order to route it correctly. Alternatively, the IXC translates the 800 number into an internal, nonstandard, 10-digit number for further routing to the terminating Central Office (CO) and trunk or trunk group.

Because 800 long distance service is essentially a database lookup and translation service for incoming phone calls, there are endless “800 services” you can create. You can put permanent instructions into the company to change the routing patterns based on time of day, day of week, number called, number calling. Some long distance companies allow you to change your routing instructions from one minute to another. For example, you might have two call centers into which 800 phone calls are pouring. When one gets busy, you may tell your long distance company to rout all the 800 inbound phone calls to the call center, which isn’t busy. See Eight Hundred Service and One Number Calling for more, especially all the features you can now get on 800 service.

In May of 1993 the FCC mandated that all 800 (and by extension all 8NN) numbers became “portable.” That means that customers can take their 800 telephone number from one long distance company to another, and still keep the same 800. See also 800 Portability.

800 services are known internationally as “Freefone Services.” In other countries the dialing scheme may vary, with examples being 0-800 and 0-500. Such services also go under the name “Greenfone.” In June 1996, the ITU-T approved the E.169 standard Universal International Freefone Number (UIFN) numbers, also known as “Global 800.” UIFN will work across national boundaries, based on an standard numbering scheme of 800, 888 or 877 plus an 8-digit telephone number. See also UIFN and Vanity Numbers.

A code assigned to a customer, a project, a department, a division—whatever.  Typically, a person dialing a long distance phone call must enter that code so the Call Accounting system can calculate and report on the cost of that call at the end of the month or designated time period.  Many service companies, such as law offices, engineering firms and advertising agencies use account codes to track costs and bill their clients accordingly.  Some account codes are very complicated.  They include the client’s number and the number of the particular project.  The Account Code then includes Client and Matter number.  These long codes can tax many call accounting systems, even some very sophisticated ones.

Asymmetric Digital Subscriber Line. One of a number of DSL technologies, and the most common one. ADSL is designed to deliver more bandwidth downstream (from the central office to the customer site) than upstream. The technical reason for this asymmetry has to do with issues of cross-coupled interference in the forms of FEXT (Far-End Crosstalk) and NEXT (Near-End Crosstalk). As it turns out, the asymmetry suits the applications perfectly, as DSL is used primarily for access to the Internet and Web, in which most people need fast downloads (music, software, presentations, etc.) but don’t need high-speed uploads. They need it for e-mail and instant messaging. In ADSL, downstream rates range from 256,000 bits per second to as much as nine million bits per second, whereas upstream bandwidth range from 16 to 640 thousand bits per second. But these figures are changing. And these days, phone companies, which are the primary providers of ADSL service, sell their offerings in all sorts of speeds. Typically the more you pay, the faster service you get. ADSL transmissions work at distances up to 18,000 feet (5,488 meters) over a single copper twisted pair. See also HDSL, SDSL, and VDSL.

ADSL was developed by Telcordia and is now standardized by ANSIU as T1.413; ETSI (European Telecommunications Standards Committee) contributed an Annex to the standards to reflect European requirements. ADSL technology splits the bandwidth of a qualifying pair to support multiple channels. An analog channel running at 4 kHz and below supports analog voice and fax. Packet data runs at 25kHz and above. Performance of ADSL lines is subject to the condition of the twisted-pair cable plant. Factors which affect performance include length of the loop, wire gauge (diameter), presence of bridge taps (better not to have any), and cross-coupled interference (NEXT and FEXT). Assuming no bridge taps and assuming 24-gauge copper,

ADSL will deliver downstream 1.5/2.0 Mbps over a distance of about 18,000 ft. (5.5 km.). At 6.1 Mbps, 12,000 ft, (3.7 km.) is the maximum length of the loop. Where the length of the loop exceeds those maximums, the achievable transmission rate drops precipitously due to signal attenuation and associated error performance. Error performance is addressed through FEC (Forward Error Correction), thereby maximizing throughput. Special electronics at both ends of the connection are required in order to accomplish the minor miracle of ADSL.  At the carrier end of the connection is placed an ATU-C (ADSL Termination Unit-Centralized), while an ATU-R (ADSL Termination Unit-Remote) is placed at the customer premises. In order to achieve such a high data rate over UTP, relatively sophisticated compression techniques must be employed. While the standard calls for us to DMT (Discrete Multi-Tone), DMT implementations have experienced some difficulty. See also ADSL Forum, ADSL Lite, ADSL2 and DSL Filter.

1. Automated Teller Machine. The street corner banking machine which is usually hooked up to a central computer through leased local lines and a multiplexed, secure data network. Some ATM networks work over ATM. See 2.

2. Asynchronous Transfer Mode. Very high speed transmission technology. ATM is a high bandwidth, low-delay, connection-oriented, packet-like switching and multiplexing technique. Usable capacity is segmented into 53-byte fixed-size cells, consisting of header and information fields, allocated to services on demand. The term “asynchronous” applies, as each cell is presented to the network on a “start-stop” basis—in other words, asynchronously. The access devices, switches and interlinking transmission facilities, of course, are all synchronized. Each ATM cell contains a 48-octet payload field, the size of which has an interesting background. Data people prefer to move data in huge blocks or frames, which are more efficient for large file transfers. Voice people, on the other hand, prefer tiny blasts of data, which are more effective for moving digitized voice samples (ala PCM in a T-Carrier environment). Since ATM is positioned as the ultimate service offering in support of data, voice data, video data, image data, and multimedia data, the small payload prevailed. With that battle out of the way, the European and U.S. camps clashed, with the European Telecommunications Standards Institute (ETSI) proposing a 32-octet cell and the U.S. Exchange Carriers Standards Association (ECSA) proposing a 64-octet cell—the issue was the difference in standard PCM voice encoding techniques. After lengthy wrangling, it was decided that a 48-octet cell would be the prefect mathematical compromise. Although neither camp was perfectly pleased (such tends to the nature of a compromise, I am told), it was a solution that all could accept. In any event, each cell also is prepended with a 5-octet Header which identifies the Virtual Path (Virtual Circuit), Virtual Channel, payload type, and cell loss priority; as well as providing for flow control, and header error control.

The small, fixed-length cells require lower processing overhead and allow higher transmission speeds than traditional packet switching methods. ATM allocates bandwidth on demand, making it suitable for high-speed connection of voice, data, and video services. ATM services will be available at access speeds up to 622 Mbps, with the backbone carrier networks operating at speeds currently as high as 2.5 Gbps. The ATM edge and core backbone switches operate at very high speeds, and typically contain multiple buses providing aggregate bandwidth of as much as 200+ Gbps. ATM core switches currently are available with capacities of as much as one terabit per second, although none have been deployed at this level.

Here’s a full explanation: Conventional networks carry data in a synchronous manner. Because empty slots are circulating even when the link is not needed, network capacity is wasted. The ATM concept which has been developed for us in broadband networks and optical fiber based systems is supported by both ITU-T (nee ITU) and ANSI standards, can also be interfaced to SONET (Synchronous Optical Network). ATM automatically adjusts the network capacity to meet the system needs and can handle data, voice, video and television signals. These are transferred in a sequence of fixed length data units called cells. Common standards definitions are provided for both private and public networks so that ATM systems can be interfaced to either or both. ATM is therefore a wideband, low delay, packet-like switching and multiplexing concept that allows flexible use of the transmission bandwidth and capable of working at data rates as high as 622.08 Mbps, with even higher [supporting document cut off abruptly].

A shortened name for an automated attendant, a device which answers a company’s phones, encourages you to touchtone in the extension you want, and rings that extension.  If that extension doesn’t answer, it may send the call to voice mail or back to the attendant.  It may also allow you to punch in digits and hear information, e.g. the company’s hours of business, addresses of local branches, etc.  See also Automated Attendant.

1. In telecommunications, bandwidth is the width of a communications channel. In analog communications, bandwidth is typically measured in Hertz—cycles per second. In digital communications, bandwidth is typically measured in bits per second (bps). A voice conversation in analog format is typically 3,000 Hertz, carried in a 4,000 Hertz analog channel. Digital communications varies depending on the method used to encode the voice. It can vary from traditional pulse code modulation at 64,000 bits per second down to voice over IP, which comes in around 10,000 bits per second, depending on the system. Do not confuse bandwidth with communications band. Let’s say we’re running a communications device in the 12 GHz band. What’s its bandwidth? That’s the space it’s occupying. Let’s say it’s occupying from 12 GHz to 12.1 GHz. This means that it’s occupying the space from 12,000,000,000 Hz to 12,100,000,000 Hz. This means its bandwidth is one hundred million cycles or one hundred megahertz (100 MHz). Affiliated terms are narrowband, wideband, and broadband. While these are not precise terms, narrowband generally refers to some number of 64 Kbps channels (Nx64) providing aggregate bandwidth less than 1.544 Mbps (24×64 Kbps, or T1), wideband is 1.544 Mbps-45 Mbps (T-1 to T-3) and broadband provides 45 Mbps (T-3) or better.

When you buy “bandwidth” from a carrier, there are three types of bandwidth you can secure:

• Effective bandwidth. The actual bandwidth that a customer receives, not the advertised bandwidth published by the service provider. The effective bandwidth is affected by noise, attenuation and other impairments.

• Headline bandwidth. What nominal bandwidth is called in the UK, New Zealand and other parts of the Commonwealth of Nations (formerly known as the British Commonwealth). Headline bandwidth is the bandwidth advertised by a broadband service provider, not the actual bandwidth that the customer receives. See effective bandwidth.

• Nominal bandwidth. The bandwidth advertised by a broadband service provider, not the actual bandwidth that the customer receives. Also called headline bandwidth.

2. The capacity to move information. A person who can master hardware, software, manufacturing and marketing—and plays the oboe or some other musical instrument—is “high bandwidth.” The term is believed to have originated in Redmond, WA, in the headquarters of Microsoft. People there (e.g. Bill Gates) who are super-intelligent and have generally broad capacities, are said to have “high bandwidth.”

3. Jargon for schedule. For example, “I have a bandwidth problem” means that I have an overloaded schedule.

4. The combined girth of a rock band. By way of example, the band “Meatloaf” is broadband, largely due to the individual girth of the singer by the same name. On the other hand, the “Rolling Stones” are narrowband, due largely to the svelte Mick Jagger. While the “Rolling Stones” are older, they are also Richard than is “Meatloaf.” So, bandwidth isn’t everything.

5. Geek term for intellect or brain power.

Also known as BRA (Basic Rate Access).  There are two standard interfaces in ISDN: BRI and PRI.  BRI is intended for consumer,SOHO(Small Office Home Office), and small business applications.  BRI supports a total signaling rate of 144 Kbps, which is divided into two B (Bearer) channels which run at 64 Kbps, and a D (Delta, or Data) channel which runs at 16 Kbps.  The B channels “bear” the actual data payload, i.e. they carry the actual information that you are sending.  Such information can be PCM-encoded digital voice, digital video, digital facsimile, or whatever you can squeeze into a 64 Kbps full-duplex channel.  The D channel is intended primarily for signaling and control information, including call setup, call maintenance and monitoring, call teardown, Caller ID, and Name ID.  As the signaling and control requirements actually are fairly modest, the D channel also will support pocket data transfer at rates up to 9.6 Kbps, by special arrangement with the servicing telephone company and at additional cost.  The preferred BRI standard is the “U” interface, which uses only two wires and makes use of the 2B1Q line coding technique in theU.S.and the 4B3T technique inEurope.  Another BRI standard is the “T” interface which uses four wires.  See ISDN for a much fuller explanation.  See also 2B1Q, 4B3T, T Interface, and U Interface.

Today’s common definition of broadband is any circuit significantly faster than a dial-up phone line.  That tends to be a cable modem circuit from your friendly local cable TV provider, a DSL circuit, a T-1 or an E-1 circuit from your friendly phone company.  In short, the term “broadband” can mean anything you want it to be so long as it’s “fast.”  In short, broadband is now more a marketing than a technical term.

Cable TeleVision.  This term originally stood for “community antenna television,” reflecting the fact that the original cable systems carried only broadcast stations received off the air; however, as cable systems began to originate their own programming, the term evolved to mean Cable Television.  CATV is a broadband transmission facility.  It generally uses a 75-ohm coaxial cable which simultaneously carries many frequency-divided TV channels.  Each channel is separated by guard channels.  Some of the industry’s first CATV pioneers were TV-set dealers who figured that cable would drive demand.  See Addressable Programming and Broadband.

Centrex is a contraction of Central Exchange. Centrex is a business telephone service offered by a local telephone company from a local central office (also called a public exchange). Centrex is basically normal single line telephone service with “bells and whistles” added. Those “bells and whistles” include intercom, call forwarding, call transfer, toll restrict, least cost routing and call hold (on single phone lines). Think about your home phone. You can often get “Custom Calling” features. These features are typically fourfold: Call Forwarding, Call Waiting, Call Conferencing and Speed Calling. Centrex is basically Custom Calling, but instead of four features, it has as many as 19 features. Like Custom Calling, Centrex features are provided by the local phone company’s central office.

Phone companies peddle Centrex which is leased to businesses as a substitute for that business buying or leasing its own on-premises telephone system—its own PBX, key system or ACD. Before Divestiture in 1984, Centrex was presumed dead. AT&T was, at that time, intent on becoming a major PBX and key system supplier. Then Divestiture came, and the operating phone companies recognized they were no longer part of AT&T, no longer had factories to support, but did have a huge number of Centrex installations providing large monthly revenues. As a result, the local operating companies injected new life into Centrex, making the service more attractive in features, price, services, and attitude. Here are the main reasons businesses go with Centrex as opposed to going with a stand-alone telephone system:

1. Money. Centrex is typically cheaper to get into (the central office already exists). Installation charges can be low. Commitment can also be low, since most Centrex service is leased on a month-to-month basis. So it’s perfect for companies planning on early move. There may be some economies of scale, also. Some phone companies are now offering low cost, large size packages.

2. Multiple locations. Companies with multiple locations in the same city often are cheaper with Centrex than with multiple private phone systems and tie lines, or with one private phone system and OPX lines. (An OPX line is an Off Premise Extension, a line going from a telephone system in one place to a phone in another. It might be used for an extension to the boss’ home.)

3. Growth. It’s theoretically easier to grow Centrex than a standalone PBX or key system, which usually has a finite limit. With Centrex, because it’s provided by a huge central office switch, it’s hard, theoretically, to run out of paths, memory, intercom lines, phones, tie lines, CO lines, etc. The limit on the growth of a Centrex is your central office, which may be thousands of lines.

4. Footprint Space Savings. You don’t have to put any switching equipment in your office. All Centrex switching equipment is at the central office. All you need at your office are phones.

5. Fewer Operators because of Centrex’s DID features. Having fewer operator positions saves money on people and space.

6. Give better service to your customers. With Centrex, each person has their own direct inward dial number. Many people prefer to dial whomever they want directly rather than going through a central operator. Saves time.

7. Better Reliability. When was the last time a central office crashed? Here are some of the features built into modern central offices: redundancy, load-sharing circuitry, power back-up, on-line diagnostics, 24-hour on-site personnel, mirror image architecture, 100% power failure phones, complete DC battery backup and battery power. Engineered to suffer fewer than three hours down time in every 40 years.

8. Non-blocking. Trunking constraints are largely eliminated with Centrex, since a central office is so large.

9. Minimal Service Costs. Repair is cheap. Service time is immediate. People are right next to the machine 24 hours a day. Phones and wires are the only things that require repair on the customers’ premises. You can easily plug new phones in and unplug them yourself. All other equipment is in the central office. You need not hold inventory or test equipment.

10. No technological obsolescence. Renting Centrex means a user has the ultimate flexibility—ability to jump quickly into new technology. Central offices are moving quickly into new technologies, such as ISDN.

11. Ability to manage it yourself. You can now get two important features previously available only on privately-owned self-contained phone systems (like PBXs): 1. The ability for you, the user, to make changes to the programming of your own Centrex installation without having to personally call a phone company representative. 2. The ability to get call detail accounting by extension and then have reports printed by a computer in your office. The phone company does this call accounting by installing a separate data line which carries Centrex call records back to the customer as those calls are made.

The above arguments are pro-Centrex. There are also anti-Centrex arguments. Central offices often run out of capacity. Centrex is also cable-intensive. A PBX with 20 trunks and 100 phones only needs 20 cable pairs from the user’s office to the telephone company. A Centrex installation with the same configuration needs 100 pairs. Every time someone new joins your company, the phone company needs to install another cable pair from the central office to your new employee’s desk. Sometimes they have it. Sometimes they don’t. Delays can get expensive. What with the explosion of telecom demand in recent years—individual fax machines, the Internet, etc. —there just isn’t enough copper in the ground, and a typical telco won’t plow in the cable unless they receive a pay-off in three years.

The “big” key to Centrex traditionally comes down to price. In some cities the price of Centrex lines is lower than “normal” PBX lines. Of course, you can buy Centrex lines and attach your own PBX or key system to those Centrex lines. The big disadvantage of Centrex is that there are very few specialized Centrex phones able to take better advantage of Centrex central office features the way electronic PBX phones take advantage of PBX features.

Centrex is known by many names among operating phone companies, including Centron and Cenpac. Centrex comes in two variations—CO and CU. CO means the Centrex service is provided by the Central Office. CU means the central office is on the customer’s premises. See the following Centrex definitions.

A 1.544 million bits per second data stream that can be configured as a single clear channel T1 interface or channelized into as many as 24 discrete DSO interfaces.  Time slots are numbered from 1 through 24.

The term and concept was coined by the Telecommunications Act of 1996.  Essentially the idea of the CLEC was that it would be a new local phone company that would compete with the incumbent, i.e. existing, monopoly local phone company.  The idea behind the Act was that the incumbent would be forced to lease local wired loops and other bits and pieces of its phone equipment—called unbundled network elements (UNE) —to the new phone company, i.e. the CLEC.  Ultimately, the theory went, the CLEC would start building its own local phone lines and installing its own equipment and the public would benefit by better, cheaper, more innovative telecom service—especially broadband service to the Internet.  The idea of leasing some of the ILEC’s plan was to give the CLEC a “leg up.”  This was the theory.  The first problem was that the legislation was the worst-written piece of legislation ever passed by Congress.  The second problem was the ILECs deeply resented the idea that they were to allow competitors to get started in business at their expense and using their equipment and their lines.  So the ILECs basically did everything they could get away with to mess up the CLECs.  That meant delaying CLECs’ orders, creating onerous, cumbersome, new rules for doing business with them and creating huge, new charges for new services.  For example, SBC (the new Southwestern Bell) came up with something called “Unbundled Local Switching” and stated in their new tariff that “The Rate Structure for ULS will be one of 2 rate structures:  Stand Alone ULS or ULS-Interim Shared Transport (ULS-IST).”  SBC laid out “General Principles for Stand Alone ULS: Stand Alone Unbundled Local Switching (ULS) which included charging for a single usage sensitive component in addition to the “appropriate” non-recurring and monthly recurring rates contained in the rate table.  No one, of course, new what any of this meant but it didn’t make any difference.  It delayed and confused things.  It was sort of like laying siege to your enemy.  And when you have unlimited resources (like the ILECs) you clearly will win.  The CLECs’ final problem was marketing and sales.  They were basically selling a service—phone or data service—that someone (their potential customers) can’t see, feel, touch or smell.  The only differentiating criterion falls on sound—the quality of which is totally indistinguishable between the CLEC and the ILEC or between the CLEC and any other phone company in the country.  The lack of a market and sales differentiator made selling CLEC services very, very difficult.  You couldn’t sell a better product, so you sold lower prices.  But no one believes the pitch for lower telecom prices.  They’ve heard that “cut price” story a thousand times since long distance was de-regulated in theU.S.in the late 1950s.  One CLEC, looking for a marketing magic bullet, did some market research among its potential customers and found that virtually all believed that the local ILEC was “The devil you know” and the local CLEC was “the devil you don’t know.”  As a result, virtually all CLECs formed in the U.S. after 1996 have essentially failed—gone bankrupt, about to go bankrupt, or are only surviving because some kindly soul is pouring good money after bad and hasn’t the guts to close down his disaster.  This may be too harsh.  There are variations on the CLEC theme that may make it, but they need to be in no way dependent on the local ILEC for anything and they need to figure in some clever way to save on their horrendously high capital expenditures and come up with some clever highly-demanded, new telecom services.  Right now, the CLECs compete on a selective basis for local phone service, long distance, international, broadband Internet access, and entertainment (e.g. Cable TV and Video on Demand).  CLECs include cellular/PCS providers, ISPs, IXCs, CATV providers, CAPs, LMDS operators and power utilities.  See Telecommunications Act of 1996 and UNE.

Another name for the Internet.  The word cloud came from the cloud symbol that you see on the computer networking charts and diagrams as representing the Internet.  See cloud computing.

Cloud computing is a general term for delivering computer services over the Internet—anything from sales force management to employee expense recording, from database management to off-site data backup, from simple email to complex scientific problems whose solutions requires seriously heavy computing power.  Cloud computing has many benefits (more about them in a moment) —so long as the cloud continues to work.  Item: I write a daily web site—www.InSearchOfThePerfectInvestment.com.  The software to create the website is in the cloud.  One day I was writing my blog when my cable TV service from Time Warner crashed, thus losing most of my morning’s work.  Time Warner did not repair my service for several hours.  In the meantime, I had to struggle to find alternate Internet service—which turned out to be Verizon’s broadband access service.  Had I not been working “in the cloud,” I would have been working on my local computer, which continued to work.  I would have had a far more productive day.  In the Spring of 2011, one of the industry’s biggest cloud services provider, Amazon (the book seller) had some major problems, which knocked many popular websites (i.e. Amazon’s customers) off the Internet.  Amazon said a traffic shift “was executed incorrectly” and went on to explain it all in a 5,700 word postmortem report.  Amazon said it will get better and hopefully the same problem won’t happen again.  Suffice, the whole area of cloud computing came from the symbol of a cloud that’s often used to represent the Internet in flow charts and diagrams.  Cloud computing services are broadly divided into three categories: Infrastructure-as-a-Service (IaaS), Platform-as-a-Service (PaaS) and Software-as-a-Service (SaaS).  A cloud computing service typically has three things that differ it from traditional computer hosting services.  It is sold on demand, typically by the minute or the hour; it is “elastic”—a user can use as much or as little as they want and the service is managed by the provider.  The customer often needs nothing more than a personal computer and Internet access.  Cloud computing has two big benefits.  First, a user doesn’t need to invest in his own computers and software and a place to house all the stuff.  Someone else has typically done all that.  That typically makes it much less expensive, especially if the user’s demand is only occasional and/or intermittent.  Second, the provider keeps the system working, i.e. maintained; and often introduces upgrades and improvements, all of which cost money.  Amazon and Google are key players in cloud computing.  Zillions of other companies are moving in.  See the following definitions.

A service provider that makes a cloud computing environment available to others.  See cloud, cloud computing, external cloud, public cloud.

(See cloud, see cloud provider, see cloud computing)

(See colocation)

A cable composed of an insulated central conducting wire wrapped in another cylindrical conducting wire.  The whole thing is usually wrapped in another insulating layer and an outer protective layer.  A coaxial cable has capacity to carry great quantities of information.  It is typically used to carry high-speed data (as in connections of 327X terminals to computer hosts) and CATV installations.

Colocation occurs when a competing local phone company (often called a CLEC, Competitive Local Exchange Carrier or Other Common Carrier) locates (i.e. puts) its switches within an incumbent local exchange carrier’s (ILEC) central office. An ILEC is the dominant phone carrier within a geographic area as determined by the F.C.C. Section 252 of the Telecommunications Act of 1996 defined Incumbent Local Exchange Carrier as a carrier that, as of the date of the enactment of the Act, provided local exchange services to a specific area. The Act provided that the Commission may treat “comparable carriers as incumbents” if they either “occupy a position in the market for telephone exchange service within an area that is comparable to the position occupied by the ILEC or such a “carrier that has substantially replaced an ILEC…” or if “such treatment is consistent with the public interest…” There are basically two types of colocation—adjacent/physical, and virtual. Adjacent and physical are the same. They mean that your equipment sits in the same building as the ILEC’s switching and cable termination equipment. Typically it sits in a locked cage. Only the CLEC and its personnel have the key. Access to that equipment is negotiated between the ILEC and the CLEC through an interconnection agreement. The CLEC will have 24×7 access to it because they have to maintain it for customers who expect services 24 hours a day, seven days a week. The concept of colocation—at this stage, a peculiarly North American idea—came about through the Telecommunications Act of 1996 U.S. It was a federal bill signed into law on February 8, 1996 “to promote competitions and reduce regulation in order to secure lower prices and higher quality services for American telecommunications consumers and encourage rapid deployment of new telecommunications technologies.” The Act required local service providers in the 100 largest metropolitan areas of the United States, the Baby Bells, to implement local Number Portability by the end of 1998. The Act also allowed the local regional Bell operating phone companies into long distance once they had met certain conditions about allowing competition in their local monopoly areas—thus the conception of collocation. That’s what the word colocation means. Okay, now to the real stuff—how to spell it. Several readers have complained that in previous editions of this dictionary it was spelled “colocation.” They point out that their non-technical English language dictionaries spell it collocation—with two “l”s. Random House Dictionary says that back in 1505-15 the word collocation appeared and was based on the Latin collocates, which derives from collocare. But Random House also includes a spelling from the era of 1965-1970, which it spells colocate and defines as to locate or be located in jointly or together, as two or more groups, military units, or the like; share or designate to share the same place. My preference is colocation, since it seems to me a logical shortening of co-location. But I’m not arguing. Choose which spelling you’d like. See also Carrier Hotel, CLEC, ELEC, ILEC, and Virtual Colocation.

Colocation also refers to the sharing of an antenna tower by two or more wireless operators. See tower colocation.

A telecommunications facility or service which permits callers from several diverse locations to be connected together for a conference call.  The conference bridge contains electronics for amplifying and balancing the conference call so everyone can hear each other and speak to each other.  The conference call’s progress is monitored through the bridge in order to produce a high quality voice conference and to maintain decent quality as people enter or leave the conference.

1. Connecting three or more people into one phone conversation. You used to have to place conference calls through an AT&T operator (you still can). But now you can also organize conference calls with most modern phone systems or a conference bridge. If conferencing is important to you, make sure your conferencing device has amplification and balancing. If not, it will simply electrically join the various conversations together and people at either end won’t be able to hear each other. There are different types of conference devices you can buy, including special teleconferencing devices that sit on conference tables and perform the function of a speakerphone, albeit a lot better. There are also dial-in devices called conference bridges. But, however you use these devices, they will require lines (and/or trunks). If you install one inside your phone system, be careful to have the extra spare extensions. For a conference of 10 people, you’ll typically need 10 extensions connected to your conference bridge. See Conference Bridge.

2. “He’s on a conference. He can’t speak with you at present.”

Optical fiber through which no light is transmitted and which, therefore, no signal is being carried.  Generally speaking, a dark fiber is one of many fibers contained within a bundle of fibers.  Carriers commonly deploy a large number of fibers (432 is a common number) at any given time, since the incremental cost of laying a big bundle is modest compared to pulling them one at a time as the need arises.  In fact, a carrier often has little choice, as the right of way may be granted once, and only once.  The fibers the carrier is using immediately are “lit,” and those that currently are unused are left “dark.”  The dark fiber is available for future use.  Sometimes dark fiber is sold by a carrier without the accompanying transmission electronics.  The customer, which may be an end user organization or another carrier, is expected to light up that strand of fiber with his own electronics.  See also dark copper, dark current, dim fiber, and lit fiber.

Communications channels provided specifically for the exchange of data as compared to voice.  See also IP Telephony.

A connection between a phone or phone system (like a PBX) and an IntereXchange Carrier (IXC) through a dedicated line. All calls over the line are automatically routed to a particular IXC.

A type of service often used by large companies which have a direct telephone line going directly to the long distance companies’ “Point of Presence” (POP), thereby bypassing the local telephone company and reducing the cost per minute.  Often referred to as “T-1” service.

A channel leased from a common carrier by an end user used exclusively by that end user.  The channel is available for use 24 hours a day, seven days a week, 52 weeks of the year, assuming it works that efficiently.

(See dedicated Channel or Circuit and see internet)

<dl>
<dt style=”margin-bottom:10px;”>Line: A nondedicated communication line in which a connection is established by dialing the destination code and then broken when the call is complete.</dt>
<dt>Switch: An Internet Access Term. Category of switching equipment designed to manage the dialup connections between the PSTN and either the Internet or a corporate LAN internetwork, providing security, accounting, and service management capabilities.</dt>
</dl>

Direct Inward Dialing.  You can dial inside a company directly without going through the attendant.  This feature used to be an exclusive feature of Centrex but it can now be provided by virtually all modern PBXs and some modern hybrids, but you must connect via specially configured DID lines from your local central office.  A DID (Direct Inward Dial) trunk is a trunk from the central office which passes the last two to four digits of the Listed Directory Number to the PBX or hybrid phone system, and the digits may then be used verbatim or modified by phone system programming to be the equivalent of an internal extension.  Therefore, an external caller may reach an internal extension by dialing a 7-digit central office number.  Notice: DID is different from a DIL (Direct-In-Line) where a standard, both-way central office trunk is programmed to always ring a specific extension or hunt group.  Traditionally, DID lines could not be used for outdial operation, since there was no dial tone offered.  More recently, the individual channels in a T-1 trunk can be defined in terms of their directional nature, with some being defined as DID, some as DOD (Direct Outward Dialing), and some as combination (both incoming and outgoing).  See also Combination Trunk, Direct Inward System Access, and DOD.

A generic name for a family of digital lines (also called xDSL) being provided by CLECs and local telephone companies to their local subscribers.  Such services go by different names and acronyms—ADSL (Asymmetric Digital Subscriber Line), HDSL (High Bit Rate Digital Subscriber Line) and SDSL (Single Pair Symmetrical Services).  Such services propose to give the subscriber up to eight million bits per second one way, downstream to the customer, and somewhat fewer bits per second upstream to the phone company.  DSL lines typically operate on one pair of wires—like a normal analog phone line.  See ADSL, G.990, G-Lite, HDSL, IDSL, RADSL, SDSL, Splitter, Splitterless, VDSL, xDSL, for more detailed explanations.

Another term for a naked DSL.  See naked DSL.

Digital Signal, level 0.  A DS-0 is a voice-grade channel of 64 Kbps (i.e. 64,000 bits per second).  This channel width is the worldwide standard speed for digitizing one voice conversation using PCM (Pulse Code Modulation).  The analog signal is sampled 8,000 times a second, with each sample encoded into an 8-bit byte (thus 8 x 8,000 = 64,000).  There are 24 DS-0 channels in a T-1, the North American version of DS-1.

Digital Signal, level 1.  It is a 1.544 Mbps (i.e. 1.544 million bits per second) in North America (T-1) and Japan (J-1), and 2.048 Mbps in Europe (E-1), i.e. 2,048,000 million bits per second.  Those speeds are symmetrical.  That means they’re the same in both directions.  Many people ask why they are different.  The first thing you have to understand is the purpose of DS-1.  The reason for developing it was to increase the number of voice grade interoffice trunks that could function over a single twisted pair of wires.  The Beta testing was conducted in New York’s Manhattan in 1961.  When the T-1 standard was developed in North America, the engineers looked at the distribution cables then in use.  What they found were cables consisting of 24 AWG wire pairs with loading coils installed every 6000 feet.  Bell Labs did some experiments and found that maximum bit rate that could be achieved over each 6000 foot span was 1.544 Mbps.  The loading coils were replaced with repeaters to terminate and re-generate the signals.  At the same time, Ericsson was following a similar path in London.  The loading coil spacing in Europe was only 4000 feet, so the maximum achievable bit rate was higher (2.048 Mpbs).  Why there’s no consistency is one of those wonderful unanswered questions.  T-1 is the original standard, having been developed by Bell Telephone Laboratories in the 1950s.  Subsequently, the ITU-T developed the European E-1 standard, variant, which also runs at 1.544 Mbps, but it is incompatible with T-1.  The T-1 standard, at 1.544 Mbps, for example, supports 24 voice conversations, each encoded at 64 Kbps.  The E-1 standard, at 2.048 Mbps, supports 30 conversations, plus two signaling and control channels, for a total of 32 channels, each of 64 Kbps.  See also T-1.

Digital Signal, level 2.  DS-2 effectively translates to T-2 in North America, and J-2 in Japan.  (There is no European equivalent.)  DS-2 supports a total signaling rate of 6.312 Mbps.  It is the equivalent of 4 T-1s, and supports 96 DS-0 channels of 64 Kbps, plus overhead in support of additional requirements for signaling and control functions.  DS-2 is used in carrier (telco) applications, and only rarely.  See DS-, DS-0, DS-1, T-1 and T-2.

Digital Signal, level 3.  In North America and Japan, DS-3 translates into T-3, which is the equivalent of 28 T-1 channels, each operating at total signaling rate of 1.544 Mbps.  The 28 T-1s are multiplexed through a M13 (Multiplex 1 to 3) Multiplexer, and 188 additional signaling and control bits are added to each T-3 frame.  As each frame is transmitted 8,000 times a second, the total T-3 signaling rate is 44.736 Mbps.  In a channelized application, T-3 supports 672 channels, each of 64 Kbps.  In the European hierarchy, a DS-3 is in the form of an E-3, which runs at a total signaling rate of 34.368 Mbps, supports 480 channels, and is the equivalent of 16 E-1s.  A J-3 runs at 32.064 Mbps, supports 480 channels, and is the equivalent of 20 J-1s.  If you’re moving a DS-3 (or any other DS signal) across continents, the standards of the target country rule.  Channels get muxed and demuxed, with signaling conventions translated, as well.  Here is a question from a reader:  On the U.S. side, T-1s are in multiples of 24×64 Kbps circuits and in the U.K. we have two-megabit circuits with 30×64 Kbps.  If we want to interconnect to the U.S. at DS-3 level, would we receive 28 T-1s with 6 spare channels, or do they muxed and demuxed into multiples of 30 when they arrive over this side of the world?  Answer: They get muxed and demuxed, along with signaling conventions translated into multiples of 30 when they arrive in the U.K.?  DS-3 is also called T-3.  See DS-, DS-0, DS-1, T-1, T-2 and T-3.

Digital Signal, level 4.  T-4 runs at a total signaling rate of 274.176 Mbps in North America in support of 168 T-1s, yielding 4032 standard voice-grade channels.  E-4 runs at 139.264 Mbps in Europe in support of 64 E-1s, yielding 1920 voice-grade channels.  J-4 runs at 397.200 Mbps in Japan in support of 240 J-1s, yielding 5760 channels.  See DS-0, DS-1, T-1, T-2 and T-3.

A T-1 line is 1.544 million bits per second transmission in both directions.  Because these 1.544 million bites are raw bits, they can be configured in many ways—as 24 voice lines or 1.544 million bit per second line to another office or the Internet.  A dynamic T-1 is not a technical term.  It’s a marketing term.  And what one vendor means by a dynamic T-1 may be very different from another vendor means.  Essentially the idea is that this T-1 is chameleon.  It changes its colors from time to time.  One moment it’s 24 voice lines.  The next moment it’s 1.544 million bits of data per second to the Internet.  Or somewhere in between, perhaps 12 lines of voice and 772,000 bits per second of data per second to and from the Internet.  In other words, it assigns bandwidth dynamically—as it’s needed.  How Dynamic T-1 works depends on your vendor, how they implement the service, and how they assign priorities between voice and data.  The typical approach is a black box at your office which monitors your T-1 traffic and reallocates bits to voice, to data, to a bit in between.  This dynamic reallocation isn’t instantaneous and in fact, may take minutes or even hours to occur.  Typically preference is given to voice, with data taking a slightly back seat.  But how dynamic T-1 is implemented varies from one vendor to another.  It pays to ask.

A Local Area Network (LAN) standard official known as IEEE 802.3 (1980)/Ethernet, and other LAN technologies are used for connecting computers, printers, workstations, terminals, servers, etc. typically within the same building or campus. That was Ethernet’s original intent, but now the Ethernet standard is embraced and used in long distance and internationally. Ethernet operates over twisted wire, coaxial cable, optical fiber and through the air (where it’s known as Wi-Fi) at speeds beginning at 10 Mbps. For LAN interconnection, Ethernet is a physical link and data link protocol reflecting the two lowest layers of the OSI Reference Model. The theoretical limit of 10-Mbps Ethernet, measured in the smallest 64-byte packets, is 14,800 pps (packets per second). By comparison, Token Ring is 30,000 and FDDI is 170,000.

Ethernet specifies a CSMA/CD (Carrier Sense Multiple Access with Collision Detection) MAC (Media Access Control) mechanism. CSMA/CD is a technique of sharing a common medium (e.g. twisted pair, or coaxial cable) among several devices. As Byte Magazine explained in its January 1991 issue, Ethernet is based on the same etiquette that makes for a polite conversation: “Listen before talking.” Of course, even when people are trying not to interrupt each other, there are those embarrassing moments when two people accidentally start talking at the same time. This is essentially what happens in Ethernet networks, where such a situation is called a “collision.” If a node on the network detects a collision, it alerts the other nodes by jamming the network with a collision “notification.” Then, after a random pause, the sending nodes try again. The messages are called frames.

Ethernet transmits variable length frames from 72 to 1518 bytes in length, each containing a header with the addresses of the source and destination stations and a trailer that contains error correction data. The preamble is 8 bytes, the destination is 6 bytes, the source is 6 bytes, the type is 2 bytes. Data is up to 1500 bytes. The final part of the Ethernet frame is the frame check sequence of 4 bytes.

The first personal computer Ethernet LAN adapter was shipped by 3Com on September 29, 1982 using the first Ethernet silicon chip from SEEQ Technology. Bob Metcalfe created the original Ethernet specification at Xerox PARC and later went on to found 3Com. In the October 31, 1994 issue of the magazine InfoWorld, Bob Metcalfe explained that Ethernet got its name “when I was writing a memo at the Xerox Palo Alto Research Center on May 22, 1973. Until then I had been calling our proposed multimegabit LAN the Alto Aloha Network. The purpose of the Alto Aloha Network was to connect experimental personal computers called Altos. And it used randomized re-transmission ideas from the University of Hawaii’s Aloha System packet radio network, circa 1970. The word either came from luminiferous ether—the omnipresent passive medium once theorized to carry electromagnetic waves through space, in particular light from the Sun to the Earth. Around the time of Einstein’s Theory of Relativity, the light-bearing ether was proven not to exist. So, in naming our LAN’s omnipresent passive medium, then a coaxial cable, which would propagate electromagnetic waves, namely data packets, I chose to recycle ether. Hence, Ethernet.”

According to Metcalfe, “Ethernet has been renamed repeatedly since 1973. In 1976, when Xerox began turning Ethernet into a product at 20 million bits per second (Mbps), we called it The Xerox Wire. When Digital, Intel, and Xerox decided in 1979 to make it a LAN standard 802.3 CSMA/CD—carrier sense multiple access with collision detection. And as the 802.3 standard evolved, it picked up such names as Thick Ethernet (IEEE 10Base-5), Thin Ethernet (10Base-2), Twisted Ethernet (10Base-T), and now Fast Ethernet (100Base-T).”

Originally, the way computers joined to an Ethernet LAN was through cards that got dropped into the computer’s bus. But quickly, with its increasing popularity, Ethernet connectivity simply got built onto the computer’s motherboard. Most computers—especially laptops, notebooks and netbooks—have at least two ways of joining to a wireless network—an RJ-45 receptacle (also called an Ethernet plug) and Wi-Fi.

The May 22, 2003 issue of my favorite magazine, the Economist, had an article on Ethernet. I excerpt it because it’s so good: “When Ethernet, now by far the most popular way of distributing data around local networks, was devised by Bob Metcalfe in a memo on May 22, 1973 at Xerox’s celebrated Palo Alto Research Centre (PARC), it was designed to send data at about three megabits per second. Today, one gigabit per second Ethernet is common and speeds of 100 gigabits per second are being developed. The vast majority of the Internet’s traffic begins and ends its journey on Ethernet networks, which are found in nearly every office network and home broadband connection. It was not supposed to be this way. Few imagined that this particular networking protocol would last as long as it has. Indeed, the landscape is littered with better-financed, better-backed rival protocols that failed against Ethernet. IBM’s Token Ring system is one famous casualty. Asynchronous Transfer Mode, supported by the telephone industry, is another. So the case of Ethernet is worth examining; the reasons for its longevity may offer lessons to the information-technology industry. Keep it simple, stupid. The first reason is simplicity. Ethernet never presupposed what sort of medium the data would travel over, be it coaxial cable or radio waves (hence the term “ether” to describe some undefined path). That made it flexible, able to incorporate improvements without challenging its fundamental design.

Second, it rapidly became an open standard at a time when most data-networking protocols were proprietary. That openness has made for a better business model. It enabled a horde of engineers from around the world to improve the technology as they competed to build inter-operable products. That competition lowered the price. What is more, the open standard meant that engineers in different organizations had to agree with each other on revised specifications, in order to avoid being cut out of the game. This ensured that the technology never became too complex or over-designed. As Charles Spurgeon, author of “Ethernet: The Definitive Guide” puts it, “It always stayed close to the ground. It addressed problems customers came up against, not problems that networking specialists thought needed to be addressed.” That, coupled with the economies of scale that come from being the entrenched technology, meant that Ethernet was faster, less expensive and less complicated to deploy than rival systems.

Third, Ethernet is based on decentralization. It lets smart “end-devices,” such as PCs, do the work of plucking the data out of the ether, rather than relying on a central unit to control the way those data are routed. In this way, Ethernet evolved in tandem with improvements in computing power-a factor that was largely overlooked by both critics and proponents when Ethernet was being pooh-poohed in the 1980s and early 1990s.

Beyond the technology, there is even a lesson for companies investing in research, albeit one learned through tears rather than triumph. Xerox failed to commercialize Ethernet, as it similarly missed exploiting other inventions created at PARC, such as the mouse and the graphical user interface. To develop Ethernet fully, Dr. Metcalfe had to leave PARC and found 3Com, now a big telecommunications-component firm. The lesson may have sunk in. In January 2002 PARC was carved out as an independent subsidiary of Xerox. That allows it to explore partnerships, spin-offs, and licensing agreements without having to get its parent’s permission.

100BaseT.  Ethernet at 100 Mbps, a tenfold improvement over the original Ethernet speed of 10 Mbps.  Fast Ethernet is in the form of an Ethernet hub with an internal bus that runs at 100 Mbps.  The interface to the hub is through a part which generally is selectable (i.e. programmable) to run at either 10 Mbps or 100 Mbps, depending on the requirement of the attached device.  Connection between the hub and the attached workstation or other device is over data-grade UTP (Unshielded Twisted Pair) in the form of Cat (Category) 5, at a minimum, and over distances of up to 100 meters, at a maximum.  The attached device connects to the UTP connection via a 10/100 Mbps NIC (Network Interface Card).  100BaseT hubs interconnect over fiber optic facilities, which can support 100 Mbps over relatively long distances with no loss of performance.  Fast Ethernet is no longer all that fast.  Gigabit Ethernet switches were standardized in 1998.  See also 10BaseT, 100BaseT, Cat 5, Ethernet, Gigabit Ethernet, NIC and UTP.

A shortened way of saying “fiber optic.”  Fiber is made of very pure glass.  In Bill Gates’ book called “The Road Ahead,” he says that “optical fiber is so clear and pure that if you looked through a wall of it 70 miles thick, you’d be able to see a candle burning on the other side.”  Digital signals, in the form of modulated light, travels on strands of fiber for long distances.  The big advantage that fiber has over copper is that it can carry far, far more information over much, much longer distances.  The short history of fiber optics for communications is that scientists keep discovering more and more ways of putting more and more information down the same one single strand of fiber.  Based on my own personal researches, no one has any idea what the eventual capacity limit of a stand of fiber optic might be.  I have personally asked many scientists (including one Nobel Physics Prize winner) and all seem to think there must be a theoretical limit.  But they don’t know what it is, or when we’ll reach that limit.  And they believe we have many, many years of breakthroughs in fiber still to go.  As of the time of this writing, SONET OC-192 (Synchronous Optical NETwork Optical Carrier Level 192) systems are being deployed fairly routinely by a number of major long distance carriers.  Each OC-192 strand supports approximately 10 Gbps.  With DWDM (Dense Wavelength Division Multiplexing), as many as 32 “windows,” or wavelengths of light, can be overlaid into a single strand at OC-192, yielding a total of approximately 320 Gbps.  Fiber is the American spelling.  The spelling in England, Europe, Canada, Australia and New Zealand is fibre.  See also the following definitions beginning with fiber.

FX.  Provides local telephone service from a central office which is outside (foreign to) the subscriber’s exchange area.  In its simplest form, a user picks up the phone in one city and receives a dial tone in the foreign city.  He will also receive calls dialed to the phone in the foreign city.  This means that people located in the foreign city can place a local call to get the user.  The airlines use a lot of foreign exchange service.  Many times, the seven digit local phone number for the airline you just called will be answered in another city, hundreds of miles away.  See also Foreign Central Office Service and Foreign Exchange Trunk.

FT-1.  Fractional T-1 refers to any data transmission rate between 56/64 Kbps (DSO rate) and 1.544 Mbps (T-1).  Fractional T-1 is a four-wire (two copper pairs) digital circuit that’s not as fast as a T-1.  Fractional T-1 is popular because it’s typically provided by a LEC (Local Exchange Carrier) or IXC (interexchange Carrier) at less cost than a full T-1, and in support of applications that don’t require the level of bandwidth provided by a full T-1.  While FT-1 is less costly than a full T-1, it is more costly on a channel-by-channel basis, as you would expect.  Users love FT-1, but carriers hate it.  FT-1 costs the carriers just as much to provision as does a full T-1, they just turn down some of the channels.  FT-1 is typically used for LAN interconnection, videoconferencing, high-speed mainframe connection and computer imaging.

Gigabit Ethernet (GE) uses the same framing as Ethernet and Fast Ethernet, but has a much higher clock speed (one billion bits per second).  There are slower Ethernets: 10Base-T Ethernet (the kind on our desktop) runs at 10 million bits per second, while fast Ethernet runs at a clock speed of one hundred million bits per second (100BaseT) —the kind we are increasingly seeing on our desktop.  The Gigabit Ethernet over fiber optic cable standard was finalized and formally approved on June 29, 1998, as IEEE 802.3z.  The GE over Category 5 cable standard was ratified as IEEE 802.3ab.  The IEEE 802.3ab standard uses all four pairs (8 wires) of cable in the Category 5 cable for transmission.  This was a departure from previous Ethernet and Fast Ethernet copper standards which only used two pairs (4 wires) in the IEEE 802.3 and 802.3u standards.  Although GE is available in both shared and switched varieties, the pricing difference between the two has become so marginal that most manufacturers are only producing the switched (and thus faster-feeling) variety.  While GE is much like traditional Ethernet, differences include frame size options.  The clock speed of GE is ten orders of magnitude greater than its predecessors and the previous frame size of 1,518 bytes was a bottleneck in the transmission of information.  The maximum frame size has been increased from 1,518 bytes to a jumbo frame size of 9K (9,216 bytes).  The larger frame size improves the frame throughput of a GE switch as each frame requires switch processing of only header information (cut-through switching).  The fewer frames presented to the switch, the more data the switch can process, switch and deliver in a given period of time.  Multi-mode fiber will support Gigabit Ethernet transmission at distances up to 550 meters, and single mode fiber up to 40 kilometers.  The distance that really can be transmitted is dependent upon the optics used to transmit and receive the signal.  In each case, there is a minimum distance of 2 meters due to issues of signal reflection (echo).  GE switches adhering to the IEEE 802.3ab standard, offer auto-negotiating 10/100/1000 Mbps ports.  Both half-duplex and full-duplex interfaces are supported, with full-duplex offering the advantage of the need for the CSMA/CD protocol because data collisions are impossible with a full-duplex mechanism.  QoS (Quality of Service) guarantees are not currently an inherent element of GE like ATM has, but Ethernet has a prioritization mechanism revolving around the three IEEE 802.1p tagging bits written about in the IEEE 802.1q standards.  Gigabit Ethernet has recently been extended to 10 gigabit speeds with the ratification of the IEEE 802.3ae.  Service providers have started to adopt 10 GE as an alternative to SONET with the advent of the IEEE 802.ah Ethernet in the First Mile (EFM) standardization.  See also 64b/66b, 802.3ah, 802.3ab, 802.3ae, and 10 Gigabit Ethernet.

Another name for IP Centrex.  See the definition of that term.

Someone else owns my PBX and rents me space, time and telecommunications services on it.  That PBX is typically away from my office.  It’s joined to me by various types of phone lines, including T-1.  Essentially hosted PBX is a fancy new name for what we used to call Centrex, when the phone company owned the PBX (actually it was a big central office, also called a public exchange) and rented me space, time and telephone services on it.  There are advantages and disadvantages to being a customer of a hosted PBX.  The arguments are similar to the age-old question: Should I buy or rent my house?  See also Centrex.

See also hosted PBX.  Also called virtual phone system.

Hosted VoIP (Voice Over Internet Protocol) is where the softswitch or IP PBX is on the service provider’s (SP) premises instead of the customer’s premises.  The SP owns the equipment and operates and maintains it.  The equipment is connected to the customer premises via fiber, DSL, T1, or even a wireless broadband link.  Hosted VoIP is analogous to hosted solutions in the POTS world, such as Centrex and hosted PBX.

An xDSL variant that uses ISDN BRI (Basic Rate Interface) technology to deliver transmission speeds of 128 Kbps on copper loops as long as 18,000 feet.  IDSL is symmetric, i.e. equal bandwidth is provided in both directions.  IDSL is a dedicated service for data communications applications only.  In that respect, IDSL differs from ISDN, which fundamentally is a circuit-switched service technology for voice, data, video and multimedia applications.  IDSL terminates at the user premise on a standard ISDN TA (Terminal Adapter).  At the LEC CO, the loops terminates in collocated ISP electronics in the form of either an IDSL access switch or an IDSL modem bank connected to a router.  The connection is then made to the ISP POP via a high-bandwidth dedicated circuit.  IDSL is used by LECs to deliver relatively low speed DSL services in geographic areas where ISDN technology is in place, but ADSL technology is not.  See also xDSL, ADSL, BRI, ISDN, HDSL, RADSL, SDSL, Terminal Adapter, and VDSL.

An AT&T term for the provision of access for multiple services such as voice and data through a single system built on common principles and providing similar service features for the different classes of services.

It’s hard to define the Internet in a way that is either meaningful nor easy to grasp. To say the Internet is the world’s largest and most complex computer and communications network is to trivialize it. But it is. To say that it is fast becoming the world’s global shopping mall (for buying and selling) is to trivialize it. To say that it’s becoming the world’s largest social network is to trivialize it. But it is. To say that it’s becoming the network for the world’s corporations to communicate is to trivialize it. But it is. To say that it’s replacing both physical mail and electronic faxes is to trivialize it. But it is. To say that it’s becoming the place for much corporate data processing to take place is to trivialize it. But it is. It clearly is the most important happening in the computing, communications, and telecommunications industries since the invention of the transistor or the computer. The Internet is both a transport network—moving every form of data around the world (voice, video, data and images) —and a network of computers which allow you (and them) to access, retrieve, process, and store all manner of information. No one really has a clear idea of the Internet’s size, its growth or its capacity. We know it grows daily and we know there are millions of web sites all over the world. The most important part of the Internet for all of us normal people is something called the World Wide Web, as characterized by all those web sites starting with www. The World Wide Web is a subset of the Internet. Web sites sit on computers worldwide joined by telecommunications links—from fiber optics to satellites. Each web site has both a number and a name, though most of us access sites only by name. Each web site’s software conforms to a specific Internet protocol called HTTP, which stands for Hypertext Transfer Protocol. HTTP is the standard way of transferring information across the Internet and the World Wide Web. The reason we can read all the documents on all the web sites is that the pages are written in HTML or Hypertext Markup Language, which tells us our Web browser (e.g. Internet Explorer, Firefox, Google Chrome, or Safari) how to display the web site and its elements (photos, videos, music, etc.). The defining feature of the Web is its ability to connect pages to one another—as well as to audio, video, and images files—with hyperlinks. Just click a link, and suddenly you’re at a web site on the other side of the world. How does the Internet find a web site like Google or Microsoft? Any computer attached at that moment to the Internet has an address—just like a phone number, except that it’s typically a 16-digit address. When you type in www.Google.com, your browser sends a command out to a database on the Internet, in effect asking, “What is www.Google.com?” Back comes the answer that it’s 173.194.35.104. Similarly for www.Microsoft.com, which is 65.55.21.250. Your browser connects you. This database is housed on a computer on the Internet called a DNS server. DNS stands for Domain Name Server. But it can get more complicated; for example, when you’re trying to reach a web site which shares one computer with several other web sites, as for example, my web site, www.TechnologyInvestor.com, does. I asked my great web hoster, ICDSoft for an explanation. Here’s their explanation:

When you try opening a website in your browser, several communication protocols are involved—TCP/IP, DNS, and HTTP. What happens is that when you enter www.technologyinvestor.com in the browser address bar, your computer sends a request to the DNS servers of your ISP for the IP address that stands for this domain. In this case it is 64.14.68.65. Then your browser connects to this IP address over TCP/IP and sends an HTTP request header for http://www.technologyinvestor.com/ (note the HTTP prefix which stands for the communication protocol). Our web server checks if there is such a site hosted on it and responds with its index page. This allows multiple sites to be hosted behind the same IP address. In case there is no such site, the server returns an error message. If you simply type the IP address in the browser, then our server will not know the contents of which website to return. An IP address and website address (URL) are separate things and there are different communication protocols involved. A website can be hosted on multiple IP addresses. Also different services can be hosted on the same IP address. For example, if you check your email for technologyinvestor.com, then the request will go to the same IP address but over a different protocol (IMAP or POP3). Communication over the Internet works over a lot of protocols on different layers, each responsible for a specific part of the connection. For example, the IP protocol ensures that the network packets will be routed to the correct location (the IP address 64.14.68.65). TCP ensures that no information will be lost and that it will be delivered at the correct speed. DNS is used for resolving a domain name to the correct IP address. HTTP is used for requesting and retrieving the correct web resource from the server. An IP address on the web is a 4- to 12-digit number. The digits are organized in four groups of numbers (which can range from 0 to 255) separated by periods.

The first question everyone asks is, “Who runs the Internet?” The simple answer is “everyone and no one.” Think of the Internet as two parts. The first are technical standards—how everyone connects to the Internet. These standards are set by various committee under the direction of something called the IETF, the Internet Engineering Task Force. The second are the communications circuits which carry the Internet’s traffic. There are hundreds of companies—including basically every traditional local and long distance phone company in the world—that interconnect. How they get paid and pay each other depends on arrangements (and arguments) they have amongst themselves.

The Internet’s roots are in a U.S. Defense Department network called Advanced Research Project Agency NETwork (ARPAnet), established in 1969. ARPAnet tied universities and research and development organizations to their military customers, and provided connectivity to a small number of supercomputer centers to support timesharing applications. Quickly, the biggest application among its users became email. Much of the funding was provided by NSFNET (National Science Foundation NETwork). In the mid-1990s, the Internet was “commercialized,” extending its use to anyone with a PC, a modem, a telephone line and an access provider—a special company known as an Internet Service Provider to Internet Access Provider. The Internet has become a major new publishing, research and commerce medium. I believe that its invention is as important to the dissemination of knowledge, to peoples’ lifestyles and to the way we’ll be conducting business in coming years as the invention of the Gutenberg Press was in 1453.

At its heart, the Internet is many large computer networks joined together over high-speed backbone data links ranging from 56 Kbps (now rare) to T-1, T-3, OC-1, OC-3 and higher. The Internet now reaches worldwide. Depending on the whim of the local government (which typically controls the local phone company and thus access to the Internet for its citizenry) you can pretty well get onto the Internet and roam it unchecked. The governments of Singapore, the People’s Republic of China, Burma, Saudi Arabia, Iran and a few others limit their peoples’ access to the Internet. They typically don’t like their citizenry reading about democracy or personal freedom. The topology of the Internet and its subnetworks changes daily, as do its providers and its content. The bottom line is that the makeup of the Internet—i.e. how it works—is not all that important. It is the applications and information available on it that are important—the most significant of which are email (electronic mail) and the World Wide Web for commerce. These days online commerce, movies, pornography and social networking have become key uses. Commercial networks from AT&T, SPRINT, Verizon, and phone companies now carry the bulk of the traffic. As NSFNET (i.e. the U.S. Government) no longer funds the Internet, it has been commercialized, with money changing hands in complex ways between users, companies with websites, Internet Access Providers, long distance providers, government, universities and others. I have no idea how many websites there are. I don’t think anyone does. It’s probably close to a billion. It’s fair to say there isn’t a company or government organization in the world that doesn’t have its own website, accessible to you and me.

The Internet’s networking technology is very smart. Every time someone hooks a new computer to the Internet, the Internet adopts that hookup as its own and begins to route Internet traffic over that hookup and through that new computer. Thus, as more computers are hooked to the Internet, its network (and its value) grows exponentially.

The Internet is basically a packet switched network based on a family of protocols called TCP/IP, which stands for Transmission Control Protocol/Internet Protocol (TCP/IP), a family of networking protocols providing communication across interconnected networks, between computers with diverse hardware architectures and between various computer operating systems. Every computer in the world, including Windows-based machines and Macintoshes, will happily communicate using TCP/IP.

How TCP Works: TCP is a reliable, connection-oriented protocol. Connection-oriented implies that TCP first establishes a connection between the two computer systems that intend to exchange data (e.g. your PC and the host computer you’re trying to reach, which may be thousands of miles away). Since most networks are built on shared media (for example, several systems sharing the same cabling), it is necessary to break chunks of data into manageable pieces so that no two communicating computers monopolize the network. These pieces are called packets. When an application sends a message to TCP for transmission, TCP breaks the message into packets, sized approximately for the network, and sends them over the network. Because a single message is often broken into many packets, TCP marks these packets with sequence numbers before sending them. The sequence numbers allow the receiving system to properly reassemble the packets into the original order, i.e. the original message. TCP checks for errors. And finally, TCP uses port IDs to specify which application running on the system is sending or receiving the data. The port ID, checksum, and sequence number are inserted into the TCP packet in a special section called the header. The header is at the beginning of the packet containing this and other “control” information for TCP.

How IP Works: IP is the messenger protocol of TCP/IP. The IP protocol, much simpler than TCP, basically addresses and sends packets. IP relies on three pieces of information, which you provide, to receive and deliver packets successfully: IP address, subnet mask, and default gateway. The IP address identifies your system on the TCP/IP network. IP addresses are 32-bit addresses that are globally unique on a network. There’s much more on TCP/IP in my definition on TCP/IP and on Internet Addresses in that definition.

Here’s how the Internet is used: As a computer network joining two (or more) computers together in a session, it is basically transparent to what it carries. It doesn’t care if it carries electronic mail, research material, shopping requests, video, images, voice phone calls, requests for information, faxes, or anything that can be digitized, placed in a packet of information and sent. A packet-switched network like the internet injects short delays into its communications as it disassembles and assembles the packets of information it sends. And while these short delays are not a problem for non-real time communications, like email, they present a problem for “real-time” information such as voice and video. The Internet can inject a delay of as much as half a second between speaking and being heard at the other end. This makes conversation difficult. Internet telephony, as it’s called when it runs on the Internet, is getting better, however, as the Internet improves and voice coding and compression techniques improve. I’ve enjoyed some relatively decent conversations to distant places.

Probably the most famous quote about the Internet is one from John Doerr, one of Silicon Valley’s most famous venture capitalists. He said, “The Internet is the greatest legal creation of wealth in the history of the planet.” Later, after the dot com bust, he came to regret his words. By hyping wealth rather than invention, he has confessed, he distracted the industry from pursuing revolutionary technologies.

Now for a little history on the Internet. In the early 1990s, the Internet was run by and for the United States Government. There was no public use of the Internet. There were no commercial applications. In fact, it wasn’t even clear to the Federal Government what the Internet actually was. So an organization called the Federal Networking Council (FNC), which actually managed networking for the Federal Government, on October 24, 1995, unanimously passed a resolution defining the term Internet. This definition was developed in consultation with the leadership of the Internet and Intellectual Property Rights (IPR) Communities. RESOLUTION:

“The Federal Networking Council (FNC) agrees that the following language reflects our definition of the term “Internet.” “Internet” refers to the global information system that:

(i) is logically linked together by a globally unique address space based on the Internet Protocol (IP) or its subsequent extensions/follow-ons;

(ii) is able to support communications using the Transmission Control Protocol/Internet Protocol (TCP/IP) suite or its subsequent extensions/follow-ons, and/or other IP-compatible protocols; and

(iii) provides, uses or makes accessible, either publicly or privately, high level services layered on the communications and related infrastructure described herein.

MCI Mail was the first commercial application attached to the Internet. Once it got on, all the other email services wanted on…and the rest is history. See the various Internet definitions following. See also Berners-Lee, Domain, Domain Naming System, Grid Computing, gTLD, ICANN, Internet2, Internet Appliance, Internet Protocol, Internet Telephony, Intranet, IP Telephony, Surf, TCP/IP, Web Browser, and Web Services.

The method by which users connect to the Internet, usually through the service of an Internet Service Provider (ISP).

The method by which users connect to the Internet, usually through the service of an Internet Service Provider (ISP).

A general term used to describe accessing the Internet using the cable TV coaxial cable for inbound Internet access (i.e. downstream) and the phone line for sending up commands and requests (i.e. upstream information).  The cable TV is very fast—as much as six million bits per second.  The phone is relatively slow—no more than fifty thousand bits per second.  But it works because most information from the Internet flows at you, not away from you.  The cable and telecom industry is working on standards to make disparate cable systems and TV set-top boxes work with each other.  The industry has developed Data Over Cable Service Interface Specification (DOCSIS), which sets standards for both two-way and cable-plus-phone specifications.  See DOCSIS.

Internet Protocol PBX.  An IP PBX connects its phones via an Ethernet LAN and sends its voice conversations in IP packets.  There are pros and cons to IP PBXs.  Move and changes with the phones are easier.  Wiring is easy.  Voice quality and management controls vary between systems.  The IP PBX is an evolving animal.  See IP Telephony and TCP/IP.  See also IP-enabled PBX.

Internet Protocol phone.  A telephone that connects to an IP PBX or VoIP provider’s equipment via an IP connection instead of via a traditional analog phone line.

See VoIP for the best explanation.  Here is Microsoft’s definition, excerpted from their white paper on TAPI 3.0: IP Telephony is an emerging set of technologies that enables voice, data, and video collaboration over existing IP-based LANs, WANs, and the Internet.  Specifically, IP Telephony uses open IETF and ITU standards to move multimedia traffic over any network that uses IP (the Internet Protocol).  This offers users both flexibility in physical media (for example, POTS lines, ADSL, ISDN, leased lines, coaxial cable, satellite, and twisted pair) and flexibility of a physical location.  As a result, the same ubiquitous networks that carry Web, email, and data traffic can be used to connect to individuals, businesses, schools, and governments worldwide.  What are the benefits of IP Telephony?  IP Telephony allows organizations and individuals to lower the costs of existing services, such as voice and broadcast video, while at the same time broadening their means of communication to include modern video conferencing, application sharing, and whiteboarding tools.  In the past, organizations have deployed separate networks to handle traditional voice, data, and video traffic.  Each with different transport requirements, these networks were expensive to install, maintain, and reconfigure.  Furthermore, since these networks were physically distinct, integration was difficult if not impossible, limiting their potential usefulness.  IP Telephony blends voice, video and data by specifying a common transport, IP, for each, effectively collapsing three networks into one.  The result is increased manageability, lower support costs, a new breed of collaboration tools, and increased productivity.  Possible applications for IP Telephony include telecommuting, real-time document collaboration, distance learning, employee training, video conferencing, video mail, and video on demand.  See the Internet, IP Telephony algorithms, TAPI, TAPI 3.0, TCP/IP, and most importantly, VoIP.

Integrated Services Digital Network. ISDN services were the telephone industry’s first attempt at providing data connection speeds for you and I at speeds faster than dial-up phone lines. With the exception of the ISDN version of T-1 (more about that in a moment), ISDN has essentially flopped at the hand of DSL, cable modems, Ethernet and the Internet. ISDN equipment cost too much, was far too difficult to connect and far too difficult to keep running. ISDN standards were set by the ITU-T (International Telecommunications Union-Telecommunications Services Sector). Those ISDN standards cover a circuit-switched digital network that supports access to any type of service (e.g. voice, data, and video) over a single, integrated local loop from the customer premises to the network edge. ISDN requires that all network elements (e.g. local loops, PBXs, and Cos) be ISDN-compatible, and that the SS7 (Signaling System 7) be in place throughout the entire network. ISDN also specifies two standard interfaces—BRI and PRI.

BRI (Basic Rate Interface) is the North American term for the low speed version known internationally as BRA (Basic Rate Access). It delivers a total of 144,000 bits per second and is designed for the desktop. (That speed is one-hundredth of what many users receive today through broadband access channels like cable modems.) BRI is known also as 2B+D. The two B (Bearer) channels are information-bearing; that is to say that they support end user data (and voice) transfer. The B channels support “clear channel” communications at 64 Kbps each. The D (Data or Delta) channel is intended primarily for signaling and control (e.g. on-hook and off-hook signaling, performance monitoring, synchronization, and error control) at 16 Kbps. The D channel also will support end user packet data transfer at speeds up to 9.6 Kbps. BRI is intended primarily for consumer and small business applications. As ISDN-compatible terminal equipment generally is too expensive, most end users had opted for a relatively inexpensive Terminal Adapter (TA) that serves as the interface between the ISDN local loop and the non-ISDN terminal equipment.

PRI (Primary Rate Interface) is the North American term for an ISDN T-1 circuit. PRI runs at a total signaling speed of 1,544,000 bits per second and supports up to 24 channels. Also known as 23B+D, PRI supports 23 Bearer channels and one D channel. Multiple PRIs can be linked to share a single D channel, as the signaling and control bandwidth requirements are relatively light; however, a backup D channel is recommended in such implementations in order to ensure that the PRI links continue to function should a D channel fail. ITU-T specifications allow as many as five PRIs to be so linked, although some manufacturers support as many as eight. The European/International version is PRA (Primary Rate Access). Also known as 30B+D, PRA supports 30 B channels and one D channel, as is the ISDN equivalent of E-1. PRI and PRA are intended for application in connecting PBXs, ACDs, and data switches, routers, and concentrators to the network. All of these various switching and concentrating devices must be ISDN-compatible. A PRI/T-1 is often installed on two unloaded copper pairs, or more commonly, on fiber.

As ISDN essentially is a highly sophisticated enhancement of the traditional circuit-switched PSTN (Public Switched Telephone Network), it offers the advantage of some flexibility. As long as ISDN is supported by all network elements at all end user locations and throughout the service provider networks, ISDN vendors pushed ISDN’s pluses, thus: First, a single local loop connecting to a single service provider can support any mix of voice, data, and video—channel-by-channel. Second, multiple channels can be linked together in what is known as “bonding” or “rate adaption.” For example, rate adaption allows you to link together two 64-KBps B channels to form a 128 Kbps chunk of bandwidth for a videoconference or, perhaps, a single symmetric (i.e. equal speed in each direction) Internet access experience at 128 Kbps. Third, ISDN is standardized worldwide (or was), so connectivity generally is not an issue. That is not to say that there are not differences from country to country or region to region, but most of those differences are relatively inconsequential at the basic level.

ISDN has “enjoyed” many “meanings,” including “I Still Don’t Know What It Means,” referring to the fact that ISDN was not well explained by the service providers; “It Still Does Nothing,” referring to the fact that ISDN does relatively little of significance (it took so long for the service provider to make it available that technologies like Frame Relay, DSL, cable modems made it obsolete); and “I Smell Dollars Now,” referring to the fact that the service providers tried to charge a lot for ISDN service.

ISDN was designed originally to be a totally new concept of what the world’s telephone system would eventually become. (Remember this “Vision” came long before the Internet.) According to AT&T, one of the original ISDN pushers, the public switched phone network had limitations:

1. Each voice line is only 4 KHz, which is very narrow, which limited also the speed you can send data across.

2. Most signaling is in-band signaling, which is very consuming of bandwidth (i.e. it’s expensive and inefficient).

3. The little out-of-band signaling that exists today runs on lines separate to the network. This includes signaling for PBX attendants, hotel/motel, Centrex and PBX calling information.

4. Most users have separate voice and data networks, which is inefficient, expensive, and limiting. (The Internet solved that one. Now everything, including voice, runs on the Internet.)

5. Premises telephone and data equipment must be separately administered from the network it runs on.

6. There is a wide and growing variety of voice, data, and digital interface standards, many of which are incompatible.

ISDN’s “vision” was to overcome these deficiencies in four ways:

1. By providing an internationally accepted standard for voice, data and signaling. That standard has pretty well achieved, though don’t try and take North American ISDN equipment to Europe.

2. By making all transmission circuits end-to-end digital.

3. By adopting a standard out-of-band signaling system.

4. By bringing significantly more bandwidth to the desktop.

One of the best features of ISDN is the speed of dialing. Instead of 20 seconds for a call to go through on today’s still partially analog network, with ISDN it takes less than a second. Also ISDN expanded the range of phone services. Here are some sample ISDN services (some of which are now available on non-ISDN phone lines): Call waiting; Citywide Centrex; Central Management of all ISDN terminals; Credit card calling; Automatic billing of certain or all calls into accounts independent of the calling line/s. Calling line identification presentation: Provides the calling party the ISDN “phone” number, possibly with additional address information, of the called party. Such information may flash across the screen of an ISDN phone or be announced by a synthesized voice. The called party can then accept, reject or transfer the call. If the called party is not there, then his/her phone will automatically record the incoming call’s phone number and allow automatic callbacks when he/she returns or calls back in from elsewhere. Calling line identification restriction; Closed user group: Restricts conversations to or among a select group of phone numbers, local, long distance or international; Collaborative Computing. Work on the same document or drawing or design with someone 10,000 miles away. Desktop videoconferencing; ISDN can carry information to and from unattended phones as long as they’re equipped with proper hardware and software. Internet Access at 128 Kbps instead of 53.3 Kbps, which is the fastest you can get on a dial-up phone line; Simultaneous Data Calls: Two users can talk and exchange information over the D packet and/or the B circuit or packet switched channel.

There were three major problems to the widespread acceptance of ISDN: First, the cost of ISDN terminal equipment was too high. Second, the cost of upgrading central office hardware to ISDN was too high. Third, Ethernet and the Internet came along. They came out of the “open” computer industry, not the “closed” telephone industry. For more on ISDN, see also AO/DI, Euro-ISDN, Intel Blue, ISDN 2, ISDN 30, ISDN Standards, ISDN Telephone, ISUP, NT1, NT2, Q.931, Robbed Bit Signaling, S Interface, SS7, SPID, T Interface, TCAP, Terminal Adapter, and U Interface.

The connection between a customer’s premises and a point of presence of the Exchange Carrier.

A local phone company.  See also LEC.  As defined by the Telecommunications Act of 1996, a local exchange carrier means any person that is engaged in the provision of telephone exchange service or exchange access.  Such term does not include a person insofar as such person is engaged in the provision of a commercial mobile service under section 332(c), except to the extent that the Commission (the Federal Communications Commission) finds that such service should be included in the definition of such term.

A telecommunications service provided by a local exchange carrier that connects a subscriber to the public switched telephone network.  Examples of local exchange service are residential single-line service (also called R1 service) and business single-line service (also called B1 service).

Any telephone call to a location outside the local service area.  Also called a toll call or trunk call.

MEN.  Metro Ethernet Network is a way to connect buildings on the Internet like desktops within a building.  Its advantages include relatively simple scalability, due to its packet-based technology.  Standards compliant interfaces are available for data communication/telecommunication devices at line rates of 10/100/1000 Mbps, and the draft standard for 10 Gbps has been ratified.  An Ethernet-based Metropolitan Area Network is generally terms a Metro Ethernet Network.  Some European service providers have also introduced MEN-like technology for Wide Area Networks.  In enterprise networks, Metro Ethernet is used primarily for two purposes: connectivity to the public Internet and connectivity between geographically separate corporate sites—an application that extends the functionality and reach ability of corporate networks.  See Metropolitan Area Network.

Multiprotocol Label Switching (MPLS) is a way for high-performance telecommunications networks to direct and carry data from one network node to the next. MPLS lets you create “virtual links” between distant nodes. It can encapsulate packets of various network protocols. In an MPLS network, data packets are assigned labels. Packet-forwarding decisions are made solely on the contents of this label, without the need to examine the packet itself. This allows you to create end-to-end circuits across any type of transport medium, using any protocol. The primary benefit is to remove dependence on a particular Data Link Layer technology, such as ATM, frame relay, SONET or Ethernet, and eliminate the need for multiple Layer 2 networks to satisfy different types of traffic. MPLS belongs to the family of packet-switched networks. MPLS operates at an OSI Model layer. It is often referred to as a “Layer 25” protocol.

It can be used to carry many different kinds of traffic, including IP packets, as well as native ATM, SONET, and Ethernet frames. MPLS works by prefixing packets with an MPLS header, containing one or more “labels.” This is called a label stack. Each label stack entry contains four fields:

• A 20-bit label value.

• A 3-bit Traffic Class field for QoS (Quality of Service) priority (experimental) and ECN (Explicit Congestion Notification).

• A 1-bit bottom of stack flag. If this is set, it signifies that the current label is the last in the stack.

• An 8-bit TTL (time to live) field.

These MPLS-labeled packets are switched after a Label Lookup/Switch. The entry and exit points of an MPLS network are called Label Edge Routers (LER), which push an MPLS label onto an incoming packet and pop it off the outgoing packet. Routers that perform routing based only on the label are called Label Switch Routers (LSR). In some applications, the packet presented to the LER already may have a label, so that the new LSR pushes a second label onto the packet.

MPLS is a family of IETF standards which Internet Protocol networks can make forwarding decisions based on a pre-allocated label to set up a Label Switched Path (LSP). MPLS grew out of Cisco’s proprietary TAG Switching protocol. MPLS has faster forwarding performance than IPv4 networks due to its ability to make decision based on the pre-allocation of a 20-bit label through the [illegible] routing protocols. MPLS works like this: As an IP data stream enters the [illegible] the network, the ingress Label Edge Router (LER) reads the destination [illegible] of the first data packet and attaches a 32-bit shim header “label” in [illegible] layer 2 and layer 3 headers of the packet. The label is mapped to [illegible] Equivalency Class (FEC) based on the destination network and the MPLS EXP value which signified the QoS level. The Label Switch Router (LSR) in the core of the network examine the 20-bit label, and switches the packet with greater speed than if the device had to interrogate the IP routing table of the device. The router swaps the label with the new label that the next router needs to assist in the completion of the LSP. There are two flavors of MPLS available: Frame-based and cell-based. Cell-based MPLS is used in ATM networks, while frame-based MPLS is used in packet-based networks like Ethernet and Frame-Relay. Although MPLS offers slight performance increases, the richness of MPLS comes from the MPLS Applications. The two MPLS applications most widely deployed are MPLS VPNs (IETF RFC2547) and MPLS Traffic Engineering.

Optical Carrier-level 12.  SONET channel of 622.08 Mbps.  See also Concatenation, OC-1, OC-12c, OC-N and SONET.

Optical Carrier-level 3.  A SONET channel equal to three DS-3s, which is equal to 155.52 Mbps.  See also Concatenation, OC-1, OC-3c, and SONET.

Optical Carrier-level 48.  SONET channel of 2.488 Gbps.  How you calculate OC-48 is to multiply 51.84 Mbps by 48.  That gives you 2.488 thousand million bits per second, or roughly 2.5 Gbps (Gigabit per second).  See also Concatenation, OC-1, OC-N and SONET.

Private Branch eXchange. A private (i.e. you, as against the phone company owns it), branch (meaning it is a small phone company central office), exchange (a central office was originally called a public exchange, or simply an exchange). In other words, a PBX is a small version of the phone company’s larger central switching office. A PBX is also called a Private Automatic Branch Exchange, though that has now become an obsolete term. In the very old days, you called the operator to make an external call, except in Europe. Then later someone made a phone system that you simply dialed nine (or another digit—in Europe it’s often zero), got a second dial tone and dialed some more digits to dial out, locally or long distance. So, the early name of Private Branch Exchange (which needed an operator) became Private AUTOMATIC Branch Exchange (which didn’t need an operator). Now, all PBXs are automatic. And now they’re all called PBXs, except overseas where they still have PBXs that are not automatic.

At the time of the Carterfone decision in the summer of 1968, PBXs were electro-mechanical step-by-step monsters. They were 100% the monopoly of the local phone company. AT&T was the major manufacturer with over 90% of all the PBXs in the U.S. GTE was next. But the Carterfone decision allowed anyone to make and sell a PBX. And the resulting inflow of manufacturers and outflow of innovation caused PBXs to go through five, six or seven generations—depending on which guru you listen to. (See my definition for GENERATIONS in this dictionary). Anyway, by the fall of 1991, PBXs were thoroughly digital, very reliable, and very full-featured. There wasn’t much you couldn’t do with them. They had oodles of features. You could combine them and make your company a mini-network. And you could buy electronic phones that made getting to all the features that much easier. Sadly, by the late 1980s, the manufacturers seemed to have finished innovating and were into price cutting. As a result, the secondary market in telephone systems was booming. Fortunately, that isn’t the end of the story. For some of the manufacturers in the late 1980s figured that if they opened their PBX’s architecture to outside computers, their customers could realize some significant benefits. (You must remember that up until this time, PBXs were one of the last remaining special purpose computers that had totally closed architecture. No one else could program them other than their makers.) Some of the benefits customers could realize from open architecture included:

—Simultaneous voice call and data screen transfer.

—Automated dial-outs from computer databases of phone numbers and automatic transfers to idle computers.

—Transfers to experts based on responses to questions, not on phone numbers.

And a million more benefits.

There are two alternatives to getting a PBX. You can buy the newer, open more full-featured version called a communications server. Or you can subscribe to your local telephone company’s Centrex service. For a long explanation on Centrex and its benefits, see Centrex. Here are some of the benefits of a PBX versus Centrex:

1. Ownership. Once you’ve paid for it, you own it. There are obvious financial and tax benefits.

2. Flexibility. A PBX is far more flexible than a central office based Centrex. A PBX has more features. You can change them faster. You can expand faster. Drop another cord in, plug some phones in, do your programming and bingo, you’re live.

3. Centrex benefits. You can always put Centrex lines behind a PBX and get the advantages of both. In some towns, Centrex lines are cheaper than PBX lines. So buy Centrex lines and put them behind your PBX. Make sure you don’t pay for Centrex features your PBX already has. (It has most.)

4. PBX phones. There are really no Centrex phones—other than a few Centrex consoles. If you want to take advantage of Centrex features, you have to punch in cumbersome, difficult-to-remember codes on typically single line phones. PBXs have electronic phones, often with screens and dedicated buttons. They’re usually a lot easier to work, a lot easier to transfer a call, conference another, etc., and a lot more productive.

5. Footprint savings. Modern PBXs take up room, more than Centrex; but the space they take up is far less than it used to be. PBXs are getting smaller.

6. Voice Processing/Automated Attendants. Centrex’s DID (Direct Inward Dialing) feature was always pushed as a big “plus.” You saved operators. However, you can now do operator-saving things with PC-based voice processing and automated attendance you couldn’t do five years ago. These things work better with on-site standalone PBXs than with distant, central office based Centrex. Moreover, virtually every PBX in existence today supports DID. You can dial directly into PBXs and reach someone at their desk just as easily as you can dial directly using Centrex.

7. Open Architecture. Most PBXs have open architecture. See OAI for the benefits. Central offices don’t.

8. Good Reliability. There have been sufficient central office crashes and sufficient improvement in the reliability of PBXs that you could happily argue that the two are on a par with each other today. Both are equally reliable, or unreliable. The only caveat, of course, is that you back your PBX up with sufficient batteries that it will last a decent power outage. Of course, that assumes that your people will be prepared to hang around and answer the phones during a blackout.

9. Expansion. Central offices are big. Allegedly you can grow your lines to whatever size you want. In contrast, PBXs have finite growth. It’s true about PBXs. But it’s equally true about central offices. I’ve personally heard too many stories about central office line shortages to believe in the nonsense about “infinite Centrex” growth. Fact is, central offices grow out, just like PBXs. Given the tight economy of recent years, local phone companies have not been buying the central offices they should have. And they have been filling central offices up a little too tight for my taste.

10. Technological obsolescence. Allegedly, central offices are upgraded faster than PBXs and therefore are always up to date technologically. It’s nonsense. The life cycle of a typical central office was 40 years until recently. It’s now around 20 years. Think of what’s happened to PCs in the past 10 years—the IBM PC debuted only in 1981—and you can imagine how obsolete many of the nation’s central offices are.

A private circuit, conversation or teleconference in which there is one person at each end, usually connected by some dedicated transmission line.  In short, a connection with only two endpoints.  See also Point-to-Multipoint.

Plain Old Telephone Service.  Pronounced POTS, like in pots and pans.  The basic service supplying standard single line telephone, telephone lines and access to the public switched network.  Nothing fancy.  No added features.  Just receive and place calls.  Nothing like Call Waiting or Call Forwarding.  They are not POTS services.  All POTS lines work on loop start signaling.  See also Loop Start.

1. Primary Rate Interface is an ISDN term used internationally to refer to what essentially is an ISDN version of a T-1 trunk from the customer premises to the edge of the ISDN service provider network. Also known as 23B+D, PRI supports 24 channels, in total. The 23B (Bearer) channels are information-bearing channels; that is to say that each supports the transfer of actual user data. Each of these B channels is a “clear channel” running at 64 Kbps. The D (Data, or Delta) channel, which also runs at 64 Kbps, is used for all signaling and control purposes (e.g. on-hook and off-hook indication, ringing signals, synchronization, performance monitoring, and error control). The D channel also can be used for end user packet data transfer when not in use for its primary function of signaling and control. Many service providers allow multiple PRIs to share a single D channel, since the signaling and control functions are not so bandwidth intensive as to require a full 64 Kilobits per second per PRI. The ITU-T, which sets ISDN standards, specifics that as many as five PRIs can share a single D channel, although a backup D channel is recommended on another PRI circuit. Therefore, the first and second PRIs each support 23B+D, and the third-fifth PRIs each support 24B+OD. The international version of PRI is known as PRA (Primary Rate Access), and supports 30B+D. See also PRA. See also ISDN for much more detail about ISDN. See also T-1.

2. Product Release Instructions. One cell phone carrier explained this term to me, “We tell our handset vendors to load our company-specific information into the handsets they deliver to us. We then add the Preferred Rooming List (PRL) before we sell the phones to our customers.” See PRL.

1. A direct circuit or channel specifically dedicated to the use of an end user organization for the purpose of directly connecting two or more sites in a multisite enterprise. A private line that connects one point to multiple points is known as point-to-multipoint. Private lines are leased from one or more carriers, which may be local or interexchange in nature. Private lines provide connectivity on a non-switched basis. As they bypass the network switches, private lines use the various switching centers (e.g. Central Offices, or Cos) only as wire centers for the interconnection of circuits. Thereby, private lines provide full-time and immediate availability, eliminating dialup delays and avoiding any potential for congestion in the core of the carrier networks.

Private lines offer highly available connectivity, as they are dedicated to the use of a single organization, which may run any type or combination of traffic types over them. As private lines are priced based on distance and bandwidth, with no usage-sensitive cost element, they can be used constantly and at maximum capacity at the same cost as if they were never used at all. Therefore, they offer a highly cost-effective to usage-sensitive, switched services in environments where communications between sites is frequent and intense. Originally, private lines were, in fact, dedicated circuits which literally could be physically traced through the network. They also were known as “nailed-up circuits,” as telephone company technicians hung the physically distinct circuits on nails driven into the walls of the central offices. Contemporary private lines actually involve dedicated channel capacity provided through the core of the carrier networks over high-capacity, multi-channel transmission facilities. The access portions (i.e. the local loop portions) of the private line are, of course, dedicated and physically distinct circuits.

Private lines are agnostic with respect to the form of the data, and the nature of the application. They can support voice, data, video, facsimile, or multimedia communications. They can run at rates of T-1, NxT-1, T-3, OC-3, or any other technically feasible speed. They can support any communications protocol, or combination of protocols, including TCP/IP, Frame Relay, or ATM. See also Private Network.

2. An outside telephone line, with a separate telephone number, which is separate from the PBX. The line is a standard business line which goes around the PBX. It connects the user directly with the LEC central office, rather than going through the PBX. Private line connections are considered to be very “private” by virtue of the fact that it is not possible for a third party (e.g. technician or console attendant) to listen to conversations without placing a physical tap on the circuit. Additionally, private lines are not subject to congestion in the PBX. As private lines also are not susceptible to catastrophic PBX failure, they often are used to provide fail-safe communications to key individuals with mission-critical responsibilities in data centers, network operations centers, and the like.

An outside telephone number separate from the PBX, can be set up to appear on one of the buttons of a key telephone.  Also called an Auxiliary Line.  See also Private Line.

Symmetrical Digital Subscriber Line, also sometimes referred to as Single-line DSL.  SDSL is a proprietary version of symmetric DSL versions such as HDSL and HDSL2.  SDSL technology offers digital bandwidth of up to 2.3 Mbps both ways (that’s why it’s called symmetrical) over a single twisted-pair copper phone line, over distances up to about 10,000 feet on an unrepeated basis.  SDSL is aimed at the corporate and SOHO markets that require high upstream and downstream traffic rates.  SDSL uses the same 2B1Q modulation scheme used in ISDN BRI.  In February 2001, the ITU-T standardized on G.shdsl, which largely obsoleted SDSL.  See also xDSL, ADSL, G.shdsl, HDSL, HDSL2, IDSL, RADSL, SHDSL, and VDSL.

1. Single Inline Package. A type of silicon chip in which all of the pins are lined up in a row. See also DIP, PGA, and SIMM.

2. SMDS Interface Protocol.

3. Session Initiation Protocol. SIP is the most important standard for setting up telephone calls, multimedia conferencing, instant messaging and other types of real-time communications on the Internet. SIP can establish sessions for features such as auto/videoconferencing, interactive gaming, and call forwarding to be deployed over IP networks, thus enabling service providers to integrate basic IP telephony services with Web, email, and chat services. SIP is much faster, more scalable and easier to implement than H.323. An array of network gear including IP phones, IP PBXs, servers, media gateways and softswitches support SIP. SIP is the Application Layer (Layer 7 of the OSI Reference Model) protocol for the establishment, modification and termination of conferencing and telephony session over an IP-based network. SIP uses text-based messages, much like HTTP. SIP was developed within the IETF MMU-SIC (Multiparty Multimedia Session Control) working group, and is defined in the IETF’s RFC 2543. SIP is touted as being much faster, more scalable and easier to implement than H.323.SIP addressing build around either a telephone number or a Web host name. In the latter case, for example, the SIP address would be based on a URL (Uniform Resource Locator), and might look something like SIP:[email protected], which makes it very easy to guess a SIP URL based on an email address. The URL is translated into an IP address through a DNS (Domain Name Server). SIP also negotiates the features and capabilities of the session at the time the session is established. For example, a caller might wish to establish a call using G.711 audio and H.261 video. The codecs embedded in the two endpoints (i.e. originating and terminating multimedia terminals) negotiate a common set of voice and video compression algorithms (which might not include G.711 and H.261), prior to establishing the session. This advance negotiation process, which relies on the Session Description Protocol (SDP), is touted as greatly reducing the call setup time required for H.323 sessions. If the called party is not available, or does not [next page of supporting document not provided] —

Session Initiation Protocol trunk. A virtual circuit set up on an Internet access line, and over which VoIP calls travel from the customer’s IP PBX to the outside world via the customer’s VoIP provider, which uses peering arrangements with other VoIP operators and/or interconnections with the PSTN to send the call to the destination network. Calls into the customer’s IP PBX from the outside world use the same interconnections, in the reverse direction. See SIP.

Synchronous Optical NETwork. A family of fiber optic transmission rates from 51.84 million bits per second to 39.812 gigabits (billion, or thousand million) per second (and going higher, as we speak), created to provide the flexibility needed to transport many digital signals with different capacities, and to provide a design standard for manufacturers.

SONET is an optical interface standard that allows interworking of transmission products from multiple vendors (i.e. mid-span meets). It defines a physical interface, optical line rates known as Optical Carrier (OC) signals, frame format and an OAM&P protocol (Operations, Administration, Maintenance, and Provisioning). The OC signals have their origins in electrical equivalents known as Synchronous Transport Signals (STSs). The base rate is 51.84 Mbps (OC-1/STS-1), which is a DS-3 (specifically, a T-3) payload of 44.736 Mbps, plus a considerable amount of overhead for network management (largely signaling and control) purposes. Higher rates are direct multiples of the base rate. Note that SONET is based in large part on T-carrier. SONET is a TDM (Time Division Multiplexed) technology, therefore, just as is T-carrier.

SONET development began at the suggestion of MCI to the Exchange Carriers Standards Association (ESCA). Bellcore then took over the project, and it ultimately came to rest at the American National Standards Institute (ANSI). Much of the development was carried out by ECDA under the auspices of ANSI. Work started on the SONET standard in the ANSI accredited T1/X1 committee in 1985, and the Phase 1 SONET standard was issued in March 1988. SONET has also been adopted by the ITU-T (International Telecommunications Union-Telecommunications Standardization Sector), previously known as the CCITT. The ITU-T version is known as SDH (Synchronous Digital Hierarchy), which varies slightly and most obviously in terms of the fact that the SDH levels begin at the OC-3 rate of 155 Mbps. In SDH, the fundamental building blocks are known as STMs (Synchronous Transport Modules) and are equivalent in rate to three SONET STS-1s. SONET is intended to attain the following goals: Multi-vendor interworking, to be cost effective for existing services on an end-to-end basis, to create an infrastructure to support new broadband services and for enhanced operations, administration, maintenance and provisioning (OAM&P). SONET offers many advantages over asynchronous transport including: Opportunity for back-to-back multiplexing, digital cross-connect panels; Easy evolution to broadband transport; Compatibility with evolving operations standards; Enhanced performance monitoring and extension of OAM&P capabilities to end users. SONET/SDH offers the critical advantage of a standard to which manufacturers can build fiber optic gear in order to ensure interconnectivity and (at least some level of) interoperability. Thereby, carriers can safely acquire and deploy multi-vendor networks without being wed to a single manufacturer. This last point was, in fact, the primary impetus for SONET development. SONET transmission equipment interleaves frames of data in simple integer multiples to form a synchronous high speed signal known as a Synchronous Transport Signal (STS). This permits easy access to low speed signals (e.g. DS-0, DS-1, etc.) without multi-stage multiplexing and demultiplexing. The low speed signals are mapped into sub-STS-1 signals called Virtual Tributaries (VTs), or Virtual Containers (VCs) in SDH. SONET uses a 51.84 MB/s STS-1 signal as the basic building block. Higher rate signals are multiples of STS-1 (e.g. the STS-12/OC-12 signal has a rate of 12 x 51.84 MB/s or 622.080 MB/s). The frame format consists of 90 x 9 bytes. The SONET frame format is divided into two main areas: Synchronous Payload Envelope (SPE) and Transport Overhead (TOH). The SPE contains the information being transported by the frame. The TOH supports the OAM&P functions of SONET, and includes a data communication channel that provides an OAM&P communication path between multiple interconnected SONET network elements. The Synchronous Payload Envelope can handle payloads in any of three ways:

1. As a continuous 50.11 Mb/s envelope for carrying asynchronous DS-3, and other payloads requiring up to 50.11 Mb/s capacity in asynchronous (byte invisible) or byte visible format;

2. In a VT (Virtual Tributary) structured enveloped to accommodate DS-1, DS-1C, DS-2, European CEPT1, or future VT-based services (see chart below). These signals can have either an asynchronous or byte visible format; and

3. As concatenated payloads to accommodate services requiring more than 50.11 Mb/s capacity. For example, three STS-1 SPEs may be concatenated to transport a broadband ISDN signal of 135 MB/s. According to AT&T, the main SONET characteristics are: A family of rates at N x 51.84 Mbps; Optical interconnect allowing mid-span meet; intraoffice mixed vendor interconnects; Overhead channels for OAM&P functions and Synchronous networking. See the table for SONET rates.

In North America, SONET rates have been limited to OC-1 plus those compatible with European SDH. Thus only OC-3, OC-12, OC-48, OC-192, and OC-768 which are equivalent to SDH-1, SDH-4, SDH-16, and SDH-64, and SDH-256, respectively; are standard.

SONET/SDH networks typically are deployed in a physical ring topology, with multiple fibers providing redundancy. In the event that a given fiber suffers a catastrophic failure, one or more other fibers are available. The rings are of two types: Line-switched and Path-Switched. SONET also may be deployed in a physical linear topology, in which case the system operates as a logical ring.

SONET/SDH has been incredibly successful in the carrier domain, although it lately has been challenged by DWDM (Dense Wavelength Division Multiplexing). The considerable advantages of SONET have been detailed above, and at some length. The criticisms of SONET include its TDM nature, which is considered inappropriate for IP traffic; its bandwidth limitations, even at 40 Gbps; and its high level of overhead, which directly reduces user data payload, although it yields considerable network management capabilities advantages. Perhaps the greatest criticism is SONET’s high cost, especially considering that an increase in bandwidth (e.g. OC-48 to OC-192) requires that the transmitting laser diode, the receiving light detector, and all intermediate optical repeaters be upgraded. DWDM is an optical transmission technique that allows multiple light signals operating at different wavelengths (i.e. frequencies of light) to share a single fiber. Thereby and for example, eight or more (eight is the point at which DWM becomes DWDM) wavelengths can each operate at 10 Gbps. As a result, DWDM offers higher aggregate speeds than SONET. DWDM also is far less expensive. On the downside, DWDM does not offer the same inherent network management capabilities and does not offer the same level of standards development, which translates into lack of interconnectivity and interoperability between network elements of disparate origin. Further, each wavelength in a DWDM system is, in essence, a separate circuit. Therefore, all traffic riding over that wavelength is transported and switched as a single entity, from point of origin to point of termination. As a result, a wavelength must carry traffic of the same type (e.g. circuit-switched voice, packet voice, IP packet data, ATM, or Frame Relay), with the same QoS (Quality of Service) requirements, original at the same place, and destined for the same place. All of that means that each wavelength must be filled to capacity, or that there must be enough available wavelengths that capacity can afford to be underutilized. The arguments over SONET vs. DWDM rage, and will continue to do so for many years. Either approach is correct, and even optimal, depending on the applications focus of a given carrier. In fact, SONET and DWDM can, and often will, coexist, with SONET-framed data riding over DWDM wavelengths. That’s my view, at least. See also ADM, DWDM, Line-Switched Ring, Path Switched Ring, SONET Interface Layers, SONET Ring, STM, Stratum Level, STS, and WDM.

Also spelled T1, which stands for Trunk Level 1. A digital transmission link with a signaling speed of 1.544 Mbps (1,544,000 bits per second) in both directions (i.e. send and receive). T-1 is a standard for digital transmission in North America—the United States and Canada. T-1 is part of a progression of digital transmission pipes—a hierarchy known generically as the DS (Digital Signal Level) hierarchy. (For the complete hierarchy, see the definition for T Carrier above.) In the olden days, T-1 was delivered to your business on two pairs of unshielded twisted copper wires—one pair for transmit and one pair for receive—the combination of these two simplex (unidirectional) circuits yields a full duplex symmetrical (bidirectional) circuit. These days, T-1 often is delivered on fiber optic lines, where fiber is available. If that’s not available, try and get it. T-1 delivered on fiber typically works better than on copper. You can lease T-1 as a channelized services (delivered as separate voice or data channels), or as an unchannelized raw bit stream (i.e. 1.536 Mbps of transmission both ways, plus .008 Mbps framing bits) and do with the 1.526 Mbps bits as you wish—the framing bits are not under your control. North American carriers typically deliver T-1 channelized, i.e. split into 23 or 24 voice-grade channels, with each running at 56/64 Kbps (i.e. 56,000 or 64,000 bits per second), depending on the generation of the channel bank equipment involved. If you have need for a bunch of local phones, it’s often cheaper to get them delivered on T-1 channels than as individual phone lines. One expensive circuit (i.e. the multi-channel T-1) is far less expensive than 24 less expensive circuits (e.g. single-channel voice circuits). While channelized T-1 was developed for and is optimized for uncompressed voice communications, it also can be used for channelized data communications. A channelized approach is required for access to the traditional PSTN, which is channelized throughout the traditional carrier networks.

On the other hand, an unchannelized approach is better for most data communications applications, and for compressed voice, video and IP telephony. The unchannelized approach provides you with 1.536 Mbps which you can split up any way you choose. If you lease a raw T-1 pipe, you could, for example, split it (i.e. multiplex it) into 12 voice grade channels to support 12 voice conversations, and use the remaining 768 Kbps for either reasonably high-speed access to the Internet or for videoconferencing with your distant office. You could also compress voice to run at speeds of perhaps 8 Kbps or less by using IP Telephony techniques and, therefore, put many more voice calls over a single T-1. Unchannelized T-1 also is commonly used for access to a frame relay or ATM network, or for Internet access. In such an application, your router or data switch or data concentrator effectively multiplexes data packets (i.e. packets, frames or cells) through the “clear” pipe. Channelization would make no sense in such an application.

In addition to use in network access applications, T-1 also can be used for private, leased line network. In a private network, you might use channelized leased T-1 PBX tie trunks to “tie” together your voice PBXs. You might use unchannelized T-1 tie trunks to directly connect your local area network routers or data switches. Note that T-1 is medium-independent. You can run it over electrical (i.e. twisted pair or coaxial cable), optical (i.e. fiber optics or infrared) or radio (i.e. microwave or satellite) transmission media.

Outside of the United States and Canada, DS-1 is called E-1, as it was developed by the CEPT (Conference of European Postal and Telecommunications Administrations) for use in Europe. E-1 runs at a total signaling rate of 2,048,000 bits per second. Only one element remains constant between it and the North American’s T-1—the DS-0, namely the 64 Kbps channel. Most often it represents a PCM voice signal sampled at 8,000 times per second, or 64,000 bits per second. However, the form of PCM encoding, also known as companding, differs between T-1 (mu-law) and E-1 (a-law). Conversion of E-1 to T-1 involves both the compression law and the signaling format. At the higher rate of 2.048 Mbps, 32 time slots are defined at the CEPT interface, but two time slots (channels) are used for non-intrusive signaling and control purposes. The remaining 30 channels are clear 64 Kbps channels for user information-voice, video, data, etc. T-1 is also called 23B+D. That means it can be channelized to contain 23 B channels and one D channel. 30B+2D is compatible with E-1—namely it can be channelized to contain 30 B channels and two D channels. See also CEPT, Channel Bank, Companding, Compression, DS-1, ISDN PRI, PCM, TDM, and Time Division Multiplexing and the following five definitions.

A telecommunications service for which the subscriber is given a toll-free number that others can call without incurring toll charges because the subscriber pays for calls to the number.

A VoIP-enabled service whereby a secondary number is established that is only able to receive incoming calls, and when such a call comes, the call is automatically forwarded over an IP network to the customer’s primary phone number.  When a virtual phone number is established for another area code, it enables long-distance charges to be avoided for calls to the virtual phone number that originates in that area code.  Similarly, when a virtual phone number is established for another country code, it means avoiding international charges.

Same as hosted phone system.  See also hosted PBX.  The phone company owns the equipment, typically located in their offices.  You get to rent (or buy) the phones sitting on your organization’s desks.

Voice over Internet Protocol. The technology used to transmit voice conversations over a data network using the Internet Protocol. Such data network may be the Internet or a corporate intranet. The “Internet Protocol” in this term is a catch-all for the protocols and technology of encoding a voice call into digital packets and then allow the voice call to be slotted in between data calls on a packet-switched data network. Such data network may be the public Internet, a corporate Intranet, or a managed network used by long distance and international communications providers. VoIP phone calls, if properly engineered, sound as good as a circuit switched TDM phone call—the ones we make and receive every day. There are three main benefits to VoIP phone calls:

First, they may potentially be cheaper. Since the data network is typically charged on a flat rate and thus the marginal cost of making a VoIP is zero, how cheaper depends on 1. The cost of terminating the VoIP call into the traditional phone network. Figure a penny a minute, or less. 2. The price of a standard circuit switched TDM call. They’ve been getting cheaper over the years. 3. How much tax is levied on both. Taxes are horrendous on traditional circuit switched long distance phone calls. They aren’t so big, yet, on VoIP calls, which are classified by some regulatory agencies as “information services,” not voice phone calls. And therefore, they escape taxes.

Second, you may achieve benefits of managing a voice and data network as one network. With IP phones, moves, adds and changes will be easier and cheaper. IP phones are basically networkable computers, each with their own addresses. The characteristic of each phone—their individual phone numbers, with their memories, their user profiles—are easily stored in a central database. Their software and their service are typically managed using standard computing systems. In short, they’re “user friendly” to manage and can be managed remotely.

Third, —and the key attraction of IP telephony—is added (and integrated) new services, including integrated messaging, voice emails, number portability, caller ID with name, call waiting, call forwarding, take your area code with you, plug into the Internet anywhere and make free calls from anywhere in the world. And best of all, you can typically manage your phone via a website on the Internet, which will tell you whish calls you made and received, etc. A VoIP phone is typically a much better animal (though also more expensive) than today’s circuit switched phone.

VoIP phones come in two flavors: first, a device that looks like a conventional phone—most often a multiline phone found in offices; second, a small “black box” which contains all the electronics to create a standard analog phone line or two out of a broadband DSL or cable modem broadband access hookup to the Internet. A residential phone user receives such a device when he subscribes to a VoIP service, for example, one costing $19.95 a month. To make phone calls, he’ll need to plug a normal analog single or multiline phone into. In fact, that’s how I make my VoIP calls.

Virtual Private Network. With a VPN, employees can log into a distant corporate local area network, server or corporate intranet over the Internet. A VPN has the look and feel of a private network to a user. But it’s really part of the Internet with heavy security—so no one on the Internet can see what’s going on in the VPN. There are several definitions for VPN, and we’ll go through them in some detail. But first, we need to explain the overall concept. A VPN is not a private network, but is virtually so, which means it’s almost so. That is to say that it exhibits at least some of the characteristics of a private network, even though it uses the resources of a public switched network. True private networks absolutely guarantee access to network resources, and security is perfect—after all, the network is a private one, comprising dedicated leased lines. Those lines (or, more commonly today, the equivalent bandwidth) have been taken out of shared public use and dedicated to the private use of an end user organization on the basis of a lease arrangement. Those dedicated leased lines often go through various switching centers (e.g. Cos or POPs), but go around, rather than through, the switches. As far as the private network is concerned, it’s a wire center, rather than a switching center. The dedicated leased lines most commonly are T-carrier or even SONET in nature, directly interconnect two or more end user sites, and can be used for any purposes the end user desires. The end user can run any higher-layer protocol it chooses—after all, it’s a private network. Sounds great, doesn’t it? Sure, it does, but the costs are high, and the complexities of designing and implementing such a network can be way of proportion to the benefits. Virtual Private Networks don’t exhibit exactly the same characteristics and, therefore, don’t perform as well as true private networks, but can come pretty close…and at much lower cost. For example, a VPN might offer priority access to bandwidth and other network resources, whereas a true private network offers guaranteed access at all times. A VPN might offer relatively tight security mechanisms, whereas a private network is totally secure. Now, let’s examine the specific definitions.

1. The first VPN was developed for voice networking, but subsequently was developed for use in data networking, as well. Also known in AT&T terminology as a Software-Defined Network (SDN), these original VPNs remain in wide use on both a domestic and an international basis. Currently, they largely are used in support of voice, as Frame Relay and other packet network technologies have proved to be more effective in support of data applications. They are a public service offered by IXCs (IntereXchange Carriers) and making use of the circuit-switched PSTN (Public Switched Telephone Network). Originally known as Switched 56, the current usage of the term “VPN” distinguishes data services offered by AT&T, MCI and Sprint from Switched 56/65 Kbps services offered by the LECs (Local Exchange Carriers, i.e. local phone companies). Although the specifics vary by IXC, VPNs offer bandwidth options of 56/64 Kbps, increments of 56/64 Kbps, 384 Kbps and 1.544 Mbps (T-1). The last two options are designed with videoconferencing in mind. VPNs provide transmission characteristics and services similar to those of private lines, including network testing, priority access, and security. Access to a circuit-switched VPN is provided over T-carrier (e.g. T-1 or Fractional T-1) local loops, which are full-duplex, four wire, digital circuits. As VPN services are dial-up services provided over the PSTN, they offer the same inherent any-to-any connectivity provided for voice calls, with the added feature of security through a Closed User Group (CUG). In other words, any location on your VPN can dial any other location on your VPN, but can’t dial any number outside the CUG and can’t be dialed by any number outside the CUG. VPNs also offer the advantage of the high level of PSTN redundancy, which translates into a high level of network resiliency. This network resiliency compares favorably to private, leased-line networks, which are highly susceptible to catastrophic failure. In fact, VPNs often are deployed as a backup to leased-line networks. VPNs also are extremely effective in support of enterprise data networking in organizations with large numbers of small sites. Small locations with relatively modest communications requirements often cannot be cost-effectively connected to long-haul, leased-line networks. VPNs offer the advantages of flexibility and scalability, as sites can be added or deleted relatively easily, with costs maintaining a fairly reasonable relationship to enterprise network functionality. The processes of network configuration (design) and reconfiguration are greatly simplified as compared to a leased-line network. Provisioning time is also greatly reduced, thanks to the flexibility of the circuit-switched network core—the only dedicated portion of the network service accessed. Compared to a private network, the greatest disadvantage of VPNs is that all calls are price based on a usage-sensitive algorithm much like that of a typical call over the PSTN. In other words, costs are calculated by duration and time of day, with prime-time calls being priced at premium. Day-of-week and other special discounted also apply. Some carriers also consider distance in the pricing of VPN calls. Note, however, that the usage-sensitive costs of a VPN typically are a lot less than the cost-per-minute of a normal dial-up call over the PSTN, sensitive to factors including the number of sites connected, usage volume commitments, and contract length. Purely from a cost standpoint, leased-lines are preferred for networking large sites with intensive communications needs. Leased line networks also can support not only data and video transmission, but also voice, thereby offering the advantage of integration of all communications needs over a single network. Access to a VPN POP (Point of Presence) can be gained directly from the IXC (IntereXchange Carrier), from a CAP (Competitive Access Provider), or from the LEC (Local Exchange Carrier). Appropriate access technologies include leased lines, Switched 56/65, and ISDN. See also Switched 56 and Private Line.

2. The second definition of VPN is a fairly generic one, referring to a packet data network service offering with some of the characteristics of a private network. Any packet data network can be used as the foundation for such a VPN, including X.25, TCP/IP, Frame Relay, and ATM networks. Each of these foundation networks is very different in terms of specifics, but they all are highly shared in terms of their basic nature. In order to provide services that emulate, or at least approximate, a private network over a highly shared network core, it is necessary to provide some additional features and mechanisms. One such feature is priority access to bandwidth, which can be accomplished through a variety of mechanisms which variously are intrinsic to the fundamental packet protocol (e.g. ATM) or through supplemental protocols (e.g. MPLS, or MultiProtocol Label Switching, which often is used in Frame Relay and TCP/IP networks). Security is a critical feature, which variously can be imposed through mechanism such as a Closed User Group (e.g. Frame Relay) or tunneling (e.g. TCP/IP).

3. In contemporary usage, VPN most commonly refers to an IP (Internet Protocol) VPN running over the public Internet. While the ubiquitous nature of the Internet is a huge advantage for data networking, the Internet is inherently both insecure and subject to variable levels of congestion. In order to create a VPN over the Internet, security issues are mitigated through the use of a combination of authentication, encryption, and tunneling. Authentication is a means of access control that confirms the identity of users through password protection or intelligent tokens, thereby reducing the possibility that unauthorized users might gain access to privileged internal computing or network resources. Authentication commonly is the responsibility of an access server running the RADIUS (Remote Access Dial-In User Service) protocol, connected to an access router with embedded firewall software. Encryption is the process of encoding, or scrambling, of the data payload prior to transmission in order to secure it; the decryption process depends on the receiver’s possession of the correct key to unlock the safety mechanism. The key is known only to the transmitting and receiving devices. Tunneling is the process of encapsulating the encrypted payload in an IP packet for secure transmission. Tunneling protocols include SOCKv5, PPTP (Point-to-Point Tunneling Protocol), L2TP (Layer 2 Tunneling Protocol), and IPSec (IP Security).

The applications scenarios for IP VPNs include remote access, intranets, and extranets. Remote access VPNs are highly effective in support of telecommuters, mobile workers, and virtual employees. Intranets are used to link branch, regional, and corporate offices. Extranets link vendors, affiliates, distributors, agents, and strategic partners into the main corporate office, with the level of access afforded being sensitive to the level of privilege indicated by a combination of password and user ID, as properly authenticated. This definition is courtesy of Ray Horak’s book, “Communications Systems and Networks.” See also Authentication, Encryption, Extranet, Firewall, Internet, Intranet, IP VPN, Tunneling, and VPN concentrator.

A service performed by Internet Service Providers (ISPs) and Internet Access Providers (IAPs) who encourage outside companies to put their websites on computers owned by the ISPs.  These computers are attached to communications links to the Internet—often high-speed links.  For this Web hosting service, the ISPs typically charge their clients by equipment and transmission capacity used.  See also Server Colocation and Web Host.

Wireless Fidelity.  Wi-Fi is now the most common way people access the Internet.  Whether at their homes, in coffee shops, in Internet cafes, in hotels, in airports, or in their offices, people open their laptops, turn on their wireless access and attach to a nearby Wi-Fi service.  Wi-Fi is a low power wireless system with a range of no more than 300 feet from the transmitter.  Hence the closer you are to a transmitter, the more change you have of connecting and the faster your Wi-Fi transmit/receive will be.  You can improve your connection chances and “extend” the distance with a device called a range extender or signal booster.  In my house, I have one installed 50 feet from my main Wi-Fi transmitter/receiver.  It allows me to move farther from my main transmitter and still get a stronger signal.  I use it for surfing the Internet in bed—an activity my wife tolerates.  Wi-Fi is defined in the IEEE’s standard 802.11b.  That standard specifies a simple low power (tops 1 WATT), unlicensed radio frequency service which actually operates on the same frequency as some cordless phones, garage door openers, walkie-talkies, etc.  An 802.11b Wi-Fi base station is typically attached to a local area network, which is then attached to the Internet and/or the corporate network through a cable modem, a DSL router, or T-1 line.  802.11b defines both the Physical (PHY) and Medium Access Control (MAC) protocols.  Specifically, the PHY spec includes three transmission options—one Ir (Infrared), and two RF (Radio Frequency).  Most Wi-Fi systems work in the 2.4 GHz range (2.4-2.483 GHz).