Sound

Sound is nothing more than the banging of air molecules together at a rapid rate. As we generate sound with our vocal chords, we are banging these air waves together to produce intelligible information that can be used and understood by others. This is in the form of air pressure changes. Hold your hand in front of your mouth as you speak. Feel the air pressure hitting your hand? This is the effect of sound as we generate the changes in air pressure. Our vocal chords, therefore, are moving back and forth and banging the molecules of air together. There is an old question, If a tree falls in the forest, and no one is there to hear it, does it create sound? The answer must be an obvious yes. Although no one is there to hear the information, the air pressure changes from the tree falling to the ground—regardless of where it falls—must still be producing sound. If we put a tape recorder in the forest and the tree falls, it is most likely that the sound will be captured on the recorder. Even though no one is there, the recorder still captures the noises created from the air changes. The telephone converts these sound waves to their analog equivalent in the form of electrical pulses, which then are carried across the telephone network. We can, therefore, assume that in order to communicate across these wires, we must have some form of sound. The sound will be converted into electricity. The electricity is then sent across the telephone wires. Sound is the banging together of air molecules at a rapid pace. This is called compression and rarefaction. Regardless, the human voice produces sound at a constantly changing set of frequencies (speed) and amplitudes Cloudness). The human voice changes these variables of frequency and amplitude in cycles per second. The vocal chords compress the air molecules at a rate of between 100 and 5000 times per second. To recreate the sound in good faith, the sound waves (or air pressure changes) are converted from sound into electricity.

This is what the telephone set is doing for us. As the electrical equivalent of sound is created, compare the sound wave to that of an electrical wave. Electricity is typically generated in an analog form by rotating the electromagnetic energy around a center point. As the energy is on the rise, it increases the amplitude to a peak level in decibels, then begins to fall. Because the wave is concurrent, it will have both a positive and a negative side of the electrical field. Therefore, as the signal decreases from the peak of the positive energy, it moves back toward the zero line. As with a magnet there are two poles, the positive and negative sides.

Therefore, as the wave gets to the zero line, it will continue to fall to the negative side of the voltage line until it hits some peak at the bottom side of the energy field. From there, it will in turn rise back up to the zero line (value) again. This, in effect, constitutes a 360° cycle around the base line, or one complete rotation. This rotation is called a sinusoidal wave. This sinusoidal wave is the analogous recreation of the human sound wave in its electrical form. This form of one wave is called one hertz, named after the gentleman who discovered this concept. Hertz is normally abbreviated as Hz. The human sound wave will have a constantly changing variable energy in both signal strength (the amplitude) and the number of rotations around the baseline over a period of time (the frequency). Voice creates these The human ear is responsive to variations of frequency and amplitude at rates from 25 to 22,000 hertz. This means that the ear can receive and discern all of the information contained in the human voice. This is important in telephony. Differences exist in human responsiveness to sound. For example, older humans can discern sounds ranging up to 7 to 8 kHz, as a result of abuses and deterioration of the eardrums.

A person who has fired a weapon (rifle or handgun) without proper protection of their ears will have damaged the upper and lower frequency responses. Whereas, a younger person who has not had the chance to damage the ears, will be able to receive and discern sound waves in the 16- to 18-kHz range. Although when we see these youngsters carrying those boomboxes on their shoulder, with the box blaring away and sitting directly next to the youngster's ear, we can only imagine how the ear will respond in a matter of time. They are inadvertently destroying the frequency responses that the ear will be able to selectively respond to. The telephone company realized that the majority of usable information in a human conversation will fit in a 3-kHz range; therefore, it provides a pair of telephone wires, typically made of copper, that will carry all of the usable information on it. The telephone networks were built to carry speech, and the most commonly carried signal on the network is in the electrical equivalent of speech (voice). The telephone transmitter converts the acoustic signal (sound wave) that is generated in the human speaker's larynx into electrical waves. Actually, the analog waves can be represented in frequency and analog changes over a broad spectrum (or band) from approximately 30 Hz to about 10 kHz. However, most of the usable and understandable energy falls in the spectrum of 200 to 3500 Hz. If we subtract the differences between the high end and the low end, the spectrum is 3300 Hz wide, or 3.3 kHz. It is not necessary to recreate all of the speech waveforms precisely to get an acceptable transmission of human speech across the telephone network. This is because the ear is not really that sensitive to very fine distinctions in the frequency changes, and the human brain can make up for any variations in the speech form by interpretation. Of course, if something does not come across the wire clearly enough, the human brain will intervene and cause the mouth to say What? This in turn will cause the transmitting end to regenerate the signal over again by repeating themselves. Because the cost of transmitting the signal across a telephone network is directly proportional to the amount of energy that must be carried, the telephone company uses a bandpass filter on the circuit. Commercially acceptable and usable information is transmitted in what is called a band-limited channel. This means that all of the usable information is allowed to pass onto the circuit, but the extraneous information that does not add significantly to the conversation is filtered off. If the extra energy is put onto the wire and carried from end to end, the costs will go up at an equal rate; this is wasteful. The human sound wave produces both frequency and amplitude changes. TOP

What Is Bandwidth?

One of the toughest concepts for anyone to understand is the discussion of bandwidth. To the novice, it becomes even more perplexing when bantered about by telephone company personnel, engineers, and others. However, it is the basis of most of what we do in the telecommunications and telephony world. Think of bandwidth as a water pipe or a garden hose. The greater the size of the pipe, the larger volume of water that will flow through the pipe. The smaller the pipe, the smaller the flow of water. Now, in telecommunications terms, think of bandwidth as a communications pipe. The bigger the pipe, the more information that will flow through it. The smaller the pipe, the less information that will be carried through the pipe. Assume, for example, that we have a lawn and we need to water it regularly to keep it green and moist. If we have an average-sized lawn, the job can be done with the standard ^ garden hose, the type that can be bought in any hardware store. When we turn on the spigot attached to our regular water pipes, a sufficient flow of water comes out of the hose. This is a regulated flow with several control mechanisms in place: The water pipes are perhaps % in diameter, allowing a certain amount of flow through them to begin with.

The garden hose is slightly smaller, so as the flow from the larger pipe to the smaller diameter garden hose occurs, the flow is restricted, but pressure allows the flow to be constant. The spigot also can be used to turn the water on full bore, or constrained to a specific flow that is more to our liking.

THE FACILITIES NETWORK The telecommunications network was a system of interconnected facilities designed to carry traffic that results from a variety of telecommunications services. When viewed from the perspective of its physical components, or facilities, the network may be referred to as the facilities network. The components of the facilities network may be divided into three broad categories.

Station equipment is generally located on the customer's premises. Its primary functions are to transmit and receive the information flow and required control signals between customers and the network.

Transmission facilities provide the communications paths that carry the information between customers. In general, transmission facilities consist of some sort of transmission medium (for example, the atmosphere, paired cable, coaxial cable, lightguide cable) and various types of electronic equipment located at different points along the transmission medium. This equipment amplifies and, sometimes, regenerates the transmitted signals. In addition, various types of facility terminal equipment provide functions needed where transmission facilities connect to switching systems and at facility junction points.

Switching Systems interconnect the transmission facilities at various key locations and route traffic through the network. The introduction of central switching into the network yields cost savings in station equipment and transmission facilities.

TRAFFIC NETWORKS The description of the telecommunications network in the previous section emphasized the physical components of the network, namely, station equipment, transmission facilities, and switching systems. In that context, the network was referred to as the facilities network. It is also important to consider the manner in which this network provides the various telecommunications; services. From this perspective, the network may be thought of as a set of traffic networks sharing common facilities. For example, the PSTN/ which provides public switched telephone network services, is the largest and best-known traffic network. Many other traffic networks provide a variety of special services such as private-line voice and data and audio and video program services. Each traffic network is designed to meet a particular set of requirements related to transmission performance, reliability, maintenance, and the ability to handle the expected traffic volume. The following sections describe some of these traffic networks and show how they use common elements and, in some cases, share them.

PUBLIC SWITCHED TELEPHONE NETWORK Because of the large volume of business and residential telephone traffic that it carries, the PSTN is probably the most familiar of the traffic networks. This network provides the public switched telephone network services. The various types of traffic in the PSTN represent communications between any two end points in the network. Traffic is switched through each switching office, or node, it encounters and travels between nodes on trunk groups. The offices and trunk groups are arranged in a hierarchical routing structure.

A TYPICAL TELEPHONE CALL To introduce the rudimentary operation of the network, this section presents a functional description of a typical telephone call, the most familiar service provided by the PSTN. The description illustrates some of the terms defined in previous sections and introduces some new terms and concepts.

SETTING THE STAGE Mrs. Cooper, a local realtor, is calling Mrs. Mahon, a prospective buyer, at her home in a neighboring town. Mrs. Cooper's telephone is served by central office A, and her central office code is 747. Mrs. Mahon's telephone is served by central office code 951 in central office B. Since many calls are placed between central offices A and B, a number of trunks provide a direct route between the two offices. An alternate route through tandem office C is also available. Direct and alternate routes for a call from Mrs. Cooper to Mrs.Mahon

INITIATING THE CALL When Mrs. Cooper picks up her handset, the switchhook contacts of the telephone set close, signaling its off-hook status. Control equipment in the switching system at office A detects a change from on-hook to off-hook status and interprets the change as a request for service. At this time, dial tone is connected to Mrs. Cooper's telephone, assuming that a register, usually called an originating register is available to accept and e the digits she will dial. After Mrs. Cooper dials the first digit, the dial tone is disconnected. digits dialed by Mrs. Cooper (951-1234) are received and stored in originating register. Initiating the call.

CALL PROCESSING AT THE ORIGINATING CENTRAL OFFICE Next, the control equipment in central office A translates the dialed number. By examining the leading digits, usually the first three, it determines that Mrs. Cooper's call is to another central office code; that is, it is not an intraoffice call. Her call is an interoffice call and must be connected to a trunk going to another office. Routing information stored in the system indicates which paths (trunk groups) are appropriate and translates the desired paths to representations of physical locations or terminations of trunks. If the call is billable, an automatic message accounting (AMA) register is requested. At this time, control equipment transfers the call information to a register in another storage area (the outpulsing register), releasing the originating register from the call. The control equipment begins scanning the outgoing trunks to find an idle trunk to office B. An idle trunk is found directly between offices A and B. The control equipment could have found that all trunks in the trunk group's) to office B were busy. In this case, it would have begun to scan the outgoing trunks to tandem switching office C, since the call could be routed on a trunk from office A to office C and from there to office B. If all trunks to tandem office C had also been busy, would have been impossible to complete the call. In that case, Mrs. cooper would have heard a reorder tone, often called a fast busy tone since it has 120 interruptions per minute (ipm), compared to the 60 ipm if the busy tone. Processing the call at the originating central office

CALL ADVANCEMENT TO THE TERMINATING CENTRAL OFFICE The first event is the seizing of an idle trunk to office B. When a trunk is seized for a particular call, it appears busy to the switching system and becomes unavailable for other calls. A equipment in Mrs. Coopers central office will periodically scan for this ready signal. When this ready signal is detected, outpulsing of digits begins. If central office B contains a single central office code, only the last four digits of Mrs. Mahon's number will be transferred. This is because all calls on the direct trunk group will terminate at central office B. However, if office B contains more than one central office code additional digits must be transmitted to identify the particular central office code serving Mrs. Mahon. Before the last digit is sent, the control equipment checks to see that the calling customer's line is still off-hook. If the calling customer has hung up (abandoned the call), the control equipment will terminate the call-processing sequence and release associated equipment and circuits. When the last digit is outpulsed, the outpulsing register is released. The digits are now stored in the incoming register at central office B. Call advancement to the terminating central office

CALL COMPLETION Once the digits are stored in an incoming register at the terminating office, many functions are initiated and supervised by the control equipment. The 4-digit line number is translated to Mrs. Mahon's physical line termination. The status of Mrs. Mahon's line is interpreted and signifies that the line is idle. (If Mrs. Mahon's line were busy, a busy signal would be returned to Mrs. Cooper.) The incoming trunk is connected through the switching network to Mrs. Mahon's line. A ringing register is seized, the incoming register is released from this call, and Mrs. Mahon's telephone rings. An audible ring, a tone that has the timing of a ringing signal and that indicates that a ringing signal is being applied to Mrs. Mahon's telephone, is sent back to Mrs. Cooper at this time. The control equipment at the terminating office will scan Mrs. Mahon's line status for an answer (off-hook) indication and, when it is detected, will terminate the ringing signal and return answer supervision to office A. This will be used to record answer or connect time for billable calls. Mrs. Mahon answers the phone, and the conversation begins. As Mrs. Cooper talks into her handset, the acoustic speech signal is converted into an electrical signal by the transmitter in the handset. The signal generated by conventional transmitters is an electrical analog of the acoustic signal. This electrical analog of the speech may proceed through the switching systems and transmission facilities to Mrs. Mahon's telephone in that form, or it may proceed through part of its path in digital form. The latter would then require analog-to-digital and digital-to-analog conversions. With conventional technology, the signal reaching Mrs. Mahon's telephone will be analog, and the receiver will convert the analog signal back to an acoustic signal. The acoustic signal from the receiver is not an exact reproduction of that at the transmitter. One reason for this is that the frequency content is limited by the transmission path. Also, impairments such as noise and loss occur, and if the call travels a long distance, an echo effect could occur. During the conversation, the originating office, office A, monitors the outgoing trunk to office B for disconnect. If the calling party hangs up first, the connection is released, and disconnect supervision is sent to the terminating office. The trunk is idled when the terminating office returns on-hook supervision. If the called party (Mrs. Mahon, in this example) hangs up first, a timed-release period of 10 to 11 seconds is initiated. The connection is released after this time—or earlier if the calling party hangs up. Completion of the call is detected and recorded at central office A for accounting purposes if there is a charge for the call; that is, if it is not covered by a fixed monthly charge or a flat rate. When the call is first dialed, the control equipment in central office A determines whether the call is billable by the routing information associated with the first three digits. If the call is billable, a register is requested from an automatic message accounting system to receive information that is to be recorded about the call. For Mrs. Cooper's call, the information recorded includes the number of Mrs. Cooper's telephone, the number dialed, the time Mrs. Mahon answered, and the time the connection was released. Data on this call and other billed calls from central office A are forwarded to a data-processing accounting center where they are periodically processed to compute customer charges. If the call is billable, Mrs. Cooper's next monthly telephone bill will include a charge for the call. Thus, a basic telecommunications service—the simple telephone call—requires a relatively complex sequence of events. TOP Completion of the call. The Telephone Network TOP A network is a series of interconnections that when all tied together form a cohesive and ubiquitous connectivity arrangement. Whew! That sounds ominous, but to make this a little simpler, look at the components of what constitutes the telephone network. Generally, a network is a series of interconnection points. The telephone companies over the years have been developing the connections throughout the country and the world, so that a level of cost-effective service can be provided to its customers. In order to build out this connectivity, the telephone companies install wires to the customer's door, whether business or residential. The spot where these wires terminate is called the demarcation point. The position of the demarcation point depends on the legal issues involved. In the beginning days of the telephone network, the telephone companies owned everything, so they ran the wires to an interface point, then connected their telephone sets to the wires at the customer's end. This was the cradle to grave service that allowed them to proliferate the connections throughout their entire operating area. New regulation in the United States, in effect since the divestiture agreement changed this connection to a point at the entrance to the customer's building. From there, the customers hook up their own equipment, items they purchase from a myriad of other sources. In the rest of the world, where full divestiture has not yet taken place, the telephone companies (or PTTs) still own the equipment. Other areas of the world have a hybrid system, where customers might or might not own their equipment. The combinations of this arrangement are almost limitless, depending on the degree of privatization and deregulation. However, the one issue that is common in most of the world to date, the local provider owns the wires from the outside world to the entrance of the customer's building. This is called the local loop. A Topology of Connections Is Used TOP At the local loop, the topological layout of the wires has traditionally been a single wire pair or multiple pairs of wires strung to the customer's location. This has always been an issue of money. Just how many pairs of wires are needed for the connection of a single line set to the network? The answer is obvious. But for other types of service, such as digital circuits and connections, the answer is two. Depending on the customer, the number of wires run to the location has been contingent on the need vs. the cost. As a result, the use of a single or dual pair of wires has been the norm. More recently, the local providers have been installing a four-pair (eight wires) connection to the customer location. This is because the customer (both business and residential) has begun to use voice lines, separate fax lines and still separate data communications hookups. Each of these require a two-wire interface, so the need for multiple pairs has grown. It is far less expensive to install multiple pairs the first time than to install a single pair of wires every time the customer asks for a new service. So, the topology is a dedicated local connection of one or more pairs from the telephone provider to the customer location. This is also called a star configuration. The telephone company connection to the customer is from a centralized point called a central office (CO). Using a star configuration, all wires home back to a centralized point, the CO. Once the hundreds or thousands of wire pairs get to the CO, things change. The provider at that point might be using a different topology. They can use a star configuration to a hierarchy of other locations in the network layout, or they can use a ring. The ring is becoming a far more prevalent method of connection for the locals. Although we might also show the ring as a triangle, it is still a functional and logical ring. This star/ring combination constitutes the bulk of the networking topologies today. At the local telephone company's (or PTTs) office, the wires are terminated in what is called a wire center. The wire center is nothing more than a very large extension of the customer's hookup. Thousands of customers come together in this centralized point. From the wire center, a series of spokes are run out to the customer direction, or to other central offices, higher level offices in the hierarchy, or wherever they need to go. The wire center is also called a frame, where all the wires are connected to the frame. At this frame, a series of cross-connections are made. These will either be to other wires that go to other locations or to a switching center where the telephone company's central computer (in older offices this can be an electromechanical system) resides. This is called the switch. Most of the equipment today is a stored-program common-controlled computer system that just happens to process cross connections for telephone calls. Remember one fundamental fact: the telephone network was designed to carry analogous electrical signals across a pair of wires to recreate a voice conversation at both ends. This network was built to carry voice. Only recently have we been transmitting other forms of communication, such as facsimile, data, and video. TOP The switch makes routing decisions based on some parameter, such as the digits dialed by the customer. As these decisions are being made very quickly, a cross connection is made in logic. This means that the switch sets up a logical connection to another set of wires. The connection can be back to the frame where the wires serve a neighboring pair of wires connected to our next-door neighbor, or to another connection that links another central office. The possibilities are only limited by the physical arrangements in the office itself.

Between and among the offices built by the carriers (the local and the long distance providers) is a set of connections usually l aid out in a ring, but it can also be a star configuration. These are called ^facilities that carry traffic.

Throughout this network, more or less connections are installed, depending on the anticipated calling patterns of the user population. Sometimes there are many connections among many offices. Other times, it can be simple and single connections.

Tied all together then is a series of local links to the customer locations, through a central office where switching and routing decisions are made, then on out to a myriad of other connections from telephone companies, long distance suppliers, and other providers. This is the basis of the telephone network.

The Local Loop

Our interface to the telephone company network is the single-line telephone set. It stands to reason that we need to connect this set to the telephone company central office (CO). The pair of twisted wires running from the telephone company's CO is called the local loop. Each subscriber, or customer, is delivered at least one pair of wires per telephone line.

There are exceptions to this rule in rural areas where the telephone company might share multiple users on a single pair of wires. This is called a party line and is again a financial decision. If the number of users demanding telephone service exceeds the number of pairs available, they might well offer the service on a party or shared set of wires. TOP

The phone company distributes its outside plant, or distribution, to the customer by running large bundles of twisted pairs toward the customer location. This is done using feeders, which are composed of 50 to over 3000 pairs of wires.

The feeders are run to splice points or breakout points called manholes or handholes. At this point, the splicing of two reels of cable will take place, assuming that the cable on a single reel was not sufficient. A lateral distribution can also take place here. Lateral distribution is the breakout of a number of pairs to run in a different direction. The lateral distribution or branch feeder is then strung to various customer locations. The end of the pair to the final customer location is called the customer pair or station drop.

It is in this outside plant, from the CO to the customer location, that 90% of all problems will occur. This is not to imply that the telco is doing a lousy job of delivering service to the customer. These cables are exposed more to cable cuts because of construction (commonly called backhoe fade), flooding at the splice locations, rodent damage, and many other risks.

In the analog dialup telephone network, each pair of the local loop is designed to carry a single telephone call to service voice conversations. This is a proven technology that works for the most part, and continues to get better as the technologies advance. The cables can be delivered via a telephone pole, buried conduit, or direct buried cables. Either way, the service is one that we are familiar with and feel comfortable with.

What has just been described is the connection at the local portion of the network. From there the local connectivity must be extended out to other locations in and around a metropolitan area, or across the country. The connections to other types of offices are then required. The Local Loop

The Network Hierarchy (Pre-1984) TOP

Prior to 1984, the network was owned by AT&T and its local Bell operating telephone companies (BOCs). It evolved through a series of interconnections based on volumes of calls and growth. A layered hierarchy of office connections was designed around a five-level architecture. Each of these layers was designed around the concept of call completion. The offices are connected together with wires of various types, called trunks. These trunks can be twisted pairs of wire, coaxial cables (like the CATV wire), radio (such as microwave), or fiberoptics. The trunks vary in their capacities, but generally high-usage trunks are used to connect between offices.

The class-5 office is the local exchange, or end office. It delivers dial tone to the customer. The end office is the closest connection to the end user. This can also have the name of a branch office. Think of a tree. The end of the branches is where all the activity takes place, and the customers are the leaves hanging off the branches. Calls between exchanges in a geographical area are connected by direct trunks between two end offices. These are called inter-office trunks. There are over 19,000 end offices in the U.S. alone that provide basic dial tone services.

The class-4 office is the toll center. A call going between two end offices that are not connected together will be routed to the class 4. The toll center is also used as the connection to the long-distance network, calls where added costs are incurred when a connection is made. This toll center could also be called a tandem office, meaning that calls would have to pass through (or tandem through) this location to get somewhere else on the network. The tandem office usually does not provide dial tone services to the end user. However, this is a variable where a single office might provide various functions. The tandem office can also be just a toll-connecting arrangement that is a pass through from various class-5 offices to the toll centers. Again, this is a variable depending on the arrangements made by the telephone providers. The ratio of toll centers that serve local long distance is approximately 9 to 1. Prior to divestiture, there were approximately 940 toll centers. TOP

The class-3 office is the primary center. Calls destined within the same state area would be passed from the local toll office to the primary center for completion. These locations are served with high-usage trunks that are used strictly for passing calls through it from one toll center to another. The primary centers would never serve dial tones to an end user. The number of primary centers prior to divestiture were approximately 170, spread across the country among the various operating telephone companies (both Bell and independent operating companies).

The class-2 office is the sectional center. A sectional center was typically the main state switching system used for interstate toll connections designated for the processing of long-distance calls from section to section. There were approximately 50+ sectional centers before the divestiture of the Bell System. These offices did not serve any end users, but would serve between primary centers in the country.

The class-1 office is the regional center. Ten regional centers existed across the country. Each center was tasked with the final set-up of calls on a region-by-region basis. However, the regional centers constitute one of the most sophisticated computer systems in the world. The regional centers continually updated each other regarding the status of every circuit in the network. These centers are required to reroute traffic around trouble spots (e.g., failed equipment or circuits) and to keep each other informed at all times. As mentioned, this was all prior to the divestiture of the local Bell operating companies by AT&T. The number of regional centers has changed to seven in the AT&T network, and might be consolidated into four mega centers in the future. Network hierarchy pre-1984

The Network Hierarchy (Post 1984)

After 1984 the network took a dramatic turn, with the separation of the Bell Operating Companies (BOCs) from AT&T. Many users screamed that things would fall apart, with service being affected. None of this came true, however. This doesn't mean that there was not a lot of confusion, there certainly was. However, things just kept humming along for the most part and calls were completed through this series of interconnecting points called the network.

The hierarchy of the network introduced a new set of terms and connections. The BOCs were classified the same as independent telephone companies. They are all called local exchange carriers (LECs). The seven spin-off companies formed as a result of the divestiture became Regional Bell Holding Companies (RBHC), which had regulated arms called the Regional Bell Operating Companies (RBOC). Each RBOC had the Bell Operating Companies in its geographical area. Additionally, the RBHCs also had an unregulated side of the business, where new ventures could be entered into (such as equipment sales, finance, real estate, etc.). TOP

Equal access, or the ability of every interexchange carrier (IEC or IXC) to connect to the Bell operating companies for long-distance service became a reality. Equal access was designed to allow the same access to other long-distance competitors that AT&T had always enjoyed prior to divestiture. Prior to divestiture, a customer attempting to use an alternative long-distance supplier would have to dial a 7- or 10-digit telephone number to get to this supplier's switch. Then, upon completion of this connection, a computer would answer the call and place a tone on the line. From there, the caller would have to enter a 7- to 11-digit authorization code. This code identified the caller by telephone number, caller name and address, and the billing arrangements. Only after the computer (switch) verified this information would it then send dial tone to the caller's ear. The caller would then have to dial the 10-digit telephone number of the requested party. This could involve very lengthy and frustrating call set-up times—especially when the called number was busy.

Users chose not to use these alternative carriers because of the time, the number of digits required, and the frustration of busy call attempts. That is, unless the organization forced the user to dial across the carriers' networks. The reason for all of the digits was simple. The telephone company did not pass on the caller information to the alternate carrier (MCI, Sprint, ITT, etc.) that they did to AT&T. Thus, the choice of many callers was AT&T because it was simpler. Now the caller information is passed on in an equal basis, hence the access is equal.

However, some of the independents and BOCs who haven't yet upgraded their offices do not connect to these IECs. These are called nonconforming end offices. The point of presence (POP) is the point where the IEC and the LEG are interconnected. This usually refers to AT&T, MCI, Sprint, and other carrier's offices. This name is now starting to change to the point of interface (POI) as opposed to POP. The main reason is that a POP implies that a switching system is at the location, which is not always true. The point of interface is where the two parties, Bell or independent, and the carrier connect. This can be a closet in a basement or hotel, connected to a dedicated trunk to another part of the country. All of this should be transparent to the end user, however. A tandem (class 4) office arrangement The network after divestiture

The North American Numbering Plan

The network numbering plan was designed to allow for the quick and discreet connection to any phone in the country. The North American numbering plan, as it is called, works on a series of 10 numbers.

The Area Code

Note that there have been some changes in this numbering plan. When it originally was formulated, the telephone numbers were divided into three sets of sequences. The first was a three-digit area code [or numbering plan assignment CNPA)]. This started with a digit from 2 to 9 in the first slot of the sequence. In the second slot, the number was set as a 0 or a 1. In the third slot, it could be any digit from 1 to 0 (0 being the 10 digit). The reasons behind this sequence were very clever. For example, the first digit did not use a 0 or 1 because these digits were used for access to operators (0) or operator services such as credit card calling, etc. The 1 was used to send a significant digit to the local switching office indicating that the call was long-distance; therefore, the switch could immediately start setting up a toll connection to the toll center or tandem office. Thus, the exclusion of 0 and 1 in the first digit of the area code facilitated the quicker call set-up. In the second slot the digit was only a 1 or 0. This was used by the switching office equipment in a screening mode. As soon as the system sampled the second digit and saw a 0 or 1, it knew that this three-digit sequence was an area code. The third slot was any digit, so it didn't matter, it was just processed normally. Back in the early 1960s, we recognized that we were running out of the area codes, given that there were only 160 available (8x2x10= 160). In reality, only 152 were allowed for use by the various states because certain ones were allotted for special services (the area codes with an Nil were always reserved, e.g., 211, 311, 411, . . . etc.). Very close administration of the area code assignments held this until 1995, when the inevitable had to occur. From the outset, the use of the entire numbering plan was limited, but it held up for over one hundred years.

The North American Numbering Plan as It Evolved

The Exchange Code

Following the area code is another three-digit sequence, called the exchange code. This is a central office designator that lists the possible number of central office codes that can be used in each area code. The exchange code originally was set up in the sequence NNX, meaning that the first and second numbers used the digits 2 through 9, for the same reasons stated in this area code. 1 and 0 were reserved for operator and long-distance access and the 0/1 exclusion in the exchange code prevented this three digit number from being confused with an area code. In the third number slot of the exchange code, any digit could be used. Clearly the greatest limitation in the exchange codes was that we would run out of exchange codes first. With NNX we have 640 possible exchange numbers to use (8 x 8 x 10 == 640), but these were used very quickly.

In the late 1960s, we began using an exchange code that changed the sequence to NXX or expanding the number of exchange codes in each of the area codes to 800. This added some relief to the numbering plan, but when using the NXX sequence the need arose for a forced 1 in advance of the 10-digit telephone number so that the call screening and number interpretation m a switch wouldn't get confused. This met with some resistance, but ultimately customers got used to the idea. The first two locations to use this revised numbering plan were Los Angeles and New York City back around early 1971.

INTERNATIONAL NUMBERING TOP

IDDD depends as directly on numbering as does national DDD, but standards are now a matter of global concern. The Comite Consultating International Telegraphique et Telephonique (CCITT) foresaw this need and organized a study to determine how to satisfy it. The standard, approved in 1964, establishes eleven digits as a preferred maximum length for international numbers but allows twelve.

The international number is flagged by a dialed prefix, not internationally standardized, that alerts the switching equipment. The international number itself consists of a country code and a national number. Country codes are standardized and vary in length from one to three digits, the first digit of which constitutes the world zone number. National numbers are the familiar telephone numbers used for domestic long-distance service.

The countries or zones anticipating the greatest telephone population by the year 2000 were assigned the shortest country codes to allow for longer national numbers. Specifically, the unified North American world zone is 1 and the Soviet Union zone is 7. In these cases, world zone and country code are the same. Other zones contain a mix of 2-digit and 3-digit country codes (for example, 52 for Mexico and 502 for Guatemala). Countries where the number of telephones to be served can be handled by nine or fewer digits are assigned 3-digit codes. Certain combinations of the initial two digits (22, 23, 24, 25, and 26 for world zone 2; 35 for world zone 3; 50 and 59 for world zone 5) are selected in forming 3-digit codes. The other pairs are assigned as 2-digit country codes. Thus, from the initial two digits, switching systems can determine whether the country code is two or three digits long. The dialing sequence for an IDDD call is illustrated by the following call from England to the United States. The customer in England dials 010-1-NXX-NXX-XXXX/ where 010 is the international subscriber dialing prefix used in England; 1 identifies North America as the world zone (and/ in this case, is the country code); and the remaining digits are the familiar 10-digit address or national number used in North America. The Bell System has authorized two prefixes for outwardbound IDDD. The prefix Oil indicates simple coin or noncoin automatic calls. The prefix 01 indicates a desire for operator assistance.

The Subscriber Extension TOP

The last sequence in the numbering plan is the subscriber extension number. This is a four-digit sequence that can use any digit in all of the slots, allotting 10,000 customer telephone numbers in each of the exchange codes. Because the four digit sequence can be composed of any numbers, the intent is to give every subscriber their own unique telephone number. This hasn't changed as yet. However, the possibility that we can still run out of numbers always exists. Therefore, a couple of ideas have been bandied around: add two, three, or four more digits to the end of the telephone numbering plan or add one or two more digits to the area code or exchange code.

In either method, users will be asked to dial more digits, an idea that is never popular, but might become a necessity. However, this is far more complicated than just adding a few more digits here and there. The whole world will be impacted by any such decision and the length of time to implement such a global change will be extraordinary.

Private Networks TOP

Many companies, depending on their size and need, create or build their own networks. These networks are usually justified on the basis of cost, availability of lines and facilities, and special need. Often, a mix of technologies are used, such as private microwave, satellite communications, fiberoptics, and infrared transmission. The initial and the ongoing costs of private ownership can be quite expensive.

Because the costs for a private network must be borne by the end user, oftentimes the telecommunications department managers are under pressure to cross subsidize the costs in other ways. Many companies with a private network have been subjected to criticisms from end users that the costs are higher than the public-switched network costs. Individual sites or departments, therefore, begin reducing their own use old the private facilities. This increases the burden on remaining users, who then foot the bill for lightly used networks.

Still others, who have built their own networks, have stated that they are amassing huge savings. When placed in a position of defending their networks, they can produce statistics of the savings reaped on a company-wide basis. If, however, they have a problem with underutilization, they can sell off some of their excess capacity.

This gives these organizations a little breathing room. Collections, maintenance, and administration costs tend to increase when reselling takes place. There is no single best answer here; each organization must look at its needs, availability, costs, and services before deciding which network to use. The private network

Hybrid Networks TOP

Some companies have to wrestle with the decision of whether to use a private or a public-switched network. This decision is not an easy one because oftentimes the numbers do not play out well. The return on their investment just will not be there.

As a result, these organizations use a mix of services based on both private and public networks. The high-end usage services, heavy volumes between two or more major company locations, will be connected via private facilities; the lower volume locations will utilize the switched network. This usually works out better financially for the organization, because the costs can be fully justified on a location-by-location basis. Pressure to install private line facilities comes from the integration of voice, data, video, graphics, and facsimile transmissions. Only by combining these services across common circuitry will many organizations realize a true savings. The hybrid network

Local Access and Transport Areas (LATAs) TOP

Local access and transport area is a term introduced with the divestiture agreement in 1984. One of the problems facing the court system was dealing with the long-distance versus local calling areas. It was a revenue sharing concern more than anything else. AT&T forced the issue by demanding that some form of revenue sharing be put in place so that a single LEG would not have the option of handling all calls and cutting out the IECs.

To solve this problem, the agreement was reached that stated the LECs would still maintain a monopoly on local dial tone and local calling. They would be restricted from carrying long-distance traffic, which would fall under the domain of the IECs. However, many states cover very large geographical areas, and calling from one end to another would be considered long distance. The country was broken down into 195 separate bounded areas for local calling (some local tolls are allowed) based on the density of population in the area. This again was a money thing. The interconnection between two LATAs would be done by the IECs.

A whole new set of acronyms emerged as a result of the divestiture agreement; LATA is only one of them. To complicate things even more, there are four types of calling using the LATA concept:

·Intrastate- belongs to the LECs

·Intrastate- belongs to the IECs

·Interstate- belongs to the IECs Interstate- can be either/or, but was originally given to the LEC

The result of this is a very confused public, and sometimes very confusing tariffs. For example, a call from Philadelphia to Los Angeles can be carried on the IEC network for as little as $0.10 per minute, yet a call from Philadelphia to Wilmington can cost as much as $0.40 per minute. The difference here is that the 3000-mile call is regulated on the long-distance basis and FCC jurisdiction. However, the 29-mile call between Philadelphia and Delaware is regulated by BOC tariff and the PUCs. These anomalies will change and are doing so now, but this is the confusion that can crop up when various players are trying to protect their interests.

Wiring Connections: Hooking Things Up TOP

The telco uses a variety of connections to bring the service to the customer locations. The typical connection is the two-wire service that we keep talking about. This two-wire interface to the network is terminated in a demarcation point as required by law. The DEMARC is the point of least penetration into the customer's premises, typically within 12 of where the telco cable comes up into the building. Normally, telco terminates in a block; this can be the standard modular block for a single-line telephone. If the customer has multiple lines, telco will terminate in a 66 block, or an RJ21X. These are fancy names for their termination points.

The typical modular connector uses an RJ11C for telephones connected to a 2-pair interface (not to be confused with the two wires), or an RJ45X as a 4-pair interface for both voice and data. Another version of connector for digital service is an 8-conductor (4-pair), called the RJ48X.

When a telco brings in a digital circuit, it will terminate the 4-wire circuit into a newer RJ68 or a smart jack. There is no major mystique in any of these connectors. The number is strictly a uniform service code so that they can keep it all straight. However, when ordering a circuit, the telco will ask you how you want it terminated. The rule of thumb, in a multiline environment, is to use the RJ21X (which is a 66 block with an amphenol connector on it). Sounds complex, doesn't it? A single line will terminate in an RJ-11C or RJ-12.

Types of Communications TOP

The next area is the directional nature of your communications channel. Three basic forms of communications channels can be selected. They are:

One way (simplex)

This is a service that is one way and only one way. You can use it to either transmit or to receive. This is not a common channel for telephony (voice) because there are very rare occasions when we speak and everyone else only listens. Feedback is one of the capabilities we pride in our communications, which would be eliminated in a one-way conversation. Broadcast television is an example of simplex communications.

Two-way alternating (half duplex)

This is the normal channel we use in a conversation. We speak to a listener, then we listen while someone else speaks. The telephone conversations we engage in are normally half duplex. Although the line is capable of handling a transmission in each direction, the human brain can't deal well with simultaneous transmit and receive.

Two-way simultaneous (duplex) or full duplex

This is used in data communications where a device can be sending to a computer and receiving from the computer at the same time. The direction of the information can be at differing speeds, such as 1200 bits/s toward the computer from the keyboard (humans cannot type much faster than that) and 9600 bits/s from the computer.

Signaling System Seven (SS7) Definition TOP

Signaling System 7 (SS7) is an architecture for performing out-of-band signaling in support of the call-establishment, billing, routing, and information-exchange functions of the public switched telephone network (PSTN). It identifies functions to be performed by a signaling-system network and a protocol to enable their performance. What is Signaling?

Signaling refers to the exchange of information between call components required to provide and maintain service.

As users of the PSTN, we exchange signaling with network elements all the time. Examples of signaling between a telephone user and the telephone network include: dialing digits, providing dial tone, accessing a voice mailbox, sending a call-waiting tone, dialing *66 (to retry a busy number), etc.

SS7 is a means by which elements of the telephone network exchange information. Information is conveyed in the form of messages. SS7 messages can convey information such as:

SS7 is characterized by high-speed packet data and out-of-band signaling. What is Out-of-Band Signaling? TOP

Out-of-band signaling is signaling that does not take place over the same path as the conversation.

We are used to thinking of signaling as being in-band. We hear dial tone, dial digits, and hear ringing over the same channel on the same pair of wires. When the call completes, we talk over the same path that was used for the signaling. Traditional telephony used to work in this way as well. The signals to set up a call between one switch and another always took place over the same trunk that would eventually carry the call. Signaling took the form of a series of multifrequency (MF) tones, much like touch tone dialing between switches.

Out-of-band signaling establishes a separate digital channel for the exchange of signaling information. This channel is called a signaling link. Signaling links are used to carry all the necessary signaling messages between nodes. Thus, when a call is placed, the dialed digits, trunk selected, and other pertinent information are sent between switches using their signaling links, rather than the trunks which will ultimately carry the conversation. Today, signaling links carry information at a rate of 56 or 64 kbps. It is interesting to note that while SS7 is used only for signaling between network elements, the ISDN D channel extends the concept of out-of-band signaling to the interface between the subscriber and the switch. With ISDN service, signaling that must be conveyed between the user station and the local switch is carried on a separate digital channel called the D channel. The voice or data which comprise the call is carried on one or more B channels.

Why Out-of-Band Signaling? TOP

Out-of-band signaling has several advantages that make it more desirable than traditional in-band signaling.

If signaling is to be carried on a different path from the voice and data traffic it supports, then what should that path look like? The simplest design would be to allocate one of the paths between each interconnected pair of switches as the signaling link. Subject to capacity constraints, all signaling traffic between the two switches could traverse this link. This type of signaling is known as associated signaling, and is shown below in Figure 1.

Figure 1. Associated Signaling

Associated signaling works well as long as a switch’s only signaling requirements are between itself and other switches to which it has trunks. If call setup and management was the only application of SS7, associated signaling would meet that need simply and efficiently. In fact, much of the out-of-band signaling deployed in Europe today uses associated mode.

The North American implementers of SS7, however, wanted to design a signaling network that would enable any node to exchange signaling with any other SS7–capable node. Clearly, associated signaling becomes much more complicated when it is used to exchange signaling between nodes which do not have a direct connection. From this need, the North American SS7 architecture was born. The North American Signaling Architecture

The North American signaling architecture defines a completely new and separate signaling network. The network is built out of the following three essential components, interconnected by signaling links:

Once deployed, the availability of SS7 network is critical to call processing. Unless SSPs can exchange signaling, they cannot complete any interswitch calls. For this reason, the SS7 network is built using a highly redundant architecture. Each individual element also must meet exacting requirements for availability. Finally, protocol has been defined between interconnected elements to facilitate the routing of signaling traffic around any difficulties that may arise in the signaling network.

To enable signaling network architectures to be easily communicated and understood, a standard set of symbols was adopted for depicting SS7 networks. Figure 2 shows the symbols that are used to depict these three key elements of any SS7 network.

Figure 2. Signaling Network Elements TOP

STPs and SCPs are customarily deployed in pairs. While elements of a pair are not generally co-located, they work redundantly to perform the same logical function. When drawing complex network diagrams, these pairs may be depicted as a single element for simplicity, as shown in Figure 3.

Figure 3. STP and SCP Pairs

Basic Signaling Architecture

Figure 4 shows a small example of how the basic elements of an SS7 network are deployed to form two interconnected networks.

Figure 4. Sample Network

The following points should be noted:

SS7 signaling links are characterized according to their use in the signaling network. Virtually all links are identical in that they are 56–kbps or 64–kbps bidirectional data links that support the same lower layers of the protocol; what is different is their use within a signaling network. The defined link types are shown in Figure 5 and defined as follows:

Figure 5. Link Types

A Links

A links interconnect an STP and either an SSP or an SCP, which are collectively referred to as signaling end points (A stands for access). A links are used for the sole purpose of delivering signaling to or from the signaling end points (they could just as well be referred to as signaling beginning points). Examples of A links are 2–8, 3–7, and 5–12 in Figure 5.

Signaling that an SSP or SCP wishes to send to any other node is sent on either of its A links to its home STP, which, in turn, processes or routes the messages. Similarly, messages intended for an SSP or SCP will be routed to one of its home STPs, which will forward them to the addressed node over its A links.

C Links TOP

C links are links that interconnect mated STPs. As will be seen later, they are used to enhance the reliability of the signaling network in instances where one or several links are unavailable. C stands for cross (7–8, 9–10, and 11–12 are C links). B links, D links, and B/D links interconnecting two mated pairs of STPs are referred to as either B links, D links, or B/D links. Regardless of their name, their function is to carry signaling messages beyond their initial point of entry to the signaling network towards their intended destination. The B stands for bridge and describes the quad of links interconnecting peer pairs of STPs. The D denotes diagonal and describes the quad of links interconnecting mated pairs of STPs at different hierarchical levels. Because there is no clear hierarchy associated with a connection between networks, interconnecting links are referred to as either B, D, or B/D links (7–11 and 7–12 are examples of B links; 8–9 and 7–10 are examples of D links; 10–13 and 9–14 are examples of interconnecting links and can be referred to as B, D, or B/D links).

E Links

While an SSP is connected to its home STP pair by a set of A links, enhanced reliability can be provided by deploying an additional set of links to a second STP pair. These links, called E (extended) links provide backup connectivity to the SS7 network in the event that the home STPs cannot be reached via the A links. While all SS7 networks include A, B/D, and C links, E links may or may not be deployed at the discretion of the network provider. The decision of whether or not to deploy E links can be made by comparing the cost of deployment with the improvement in reliability. (1–11 and 1–12 are E links.)

F Links TOP

F (fully associated) links are links which directly connect two signaling end points. F links allow associated signaling only. Because they bypass the security features provided by an STP, F links are not generally deployed between networks. Their use within an individual network is at the discretion of the network provider. (1–2 is an F link.) Basic Call Setup Example

Before going into much more detail, it might be helpful to look at several basic calls and the way in which they use SS7 signaling (see Figure 6).

Figure 6. Call Setup Example

In this example, a subscriber on switch A places a call to a subscriber on switch B.

People generally are familiar with the toll-free aspect of 800 (or 888) numbers, but these numbers have significant additional capabilities made possible by the SS7 network. 800 numbers are virtual telephone numbers. Although they are used to point to real telephone numbers, they are not assigned to the subscriber line itself.

When a subscriber dials an 800 number, it is a signal to the switch to suspend the call and seek further instructions from a database. The database will provide either a real phone number to which the call should be directed, or it will identify another network (e.g., a long-distance carrier) to which the call should be routed for further processing. While the response from the database could be the same for every call (as, for example, if you have a personal 800 number), it can be made to vary based on the calling number, the time of day, the day of the week, or a number of other factors.

The following example shows how an 800 call is routed (see Figure 7).

Figure 7. Database Query Example

As the call-flow examples show, the SS7 network is an interconnected set of network elements that is used to exchange messages in support of telecommunications functions. The SS7 protocol is designed to both facilitate these functions and to maintain the network over which they are provided. Like most modern protocols, the SS7 protocol is layered.

Physical Layer TOP

This defines the physical and electrical characteristics of the signaling links of the SS7 network. Signaling links utilize DS–0 channels and carry raw signaling data at a rate of 56 kbps or 64 kbps (56 kbps is the more common implementation).

Message Transfer Part—Level 2

The level 2 portion of the message transfer part (MTP Level 2) provides link-layer functionality. It ensures that the two end points of a signaling link can reliably exchange signaling messages. It incorporates such capabilities as error checking, flow control, and sequence checking.

Message Transfer Part—Level 3

The level 3 portion of the message transfer part (MTP Level 3) extends the functionality provided by MTP level 2 to provide network layer functionality. It ensures that messages can be delivered between signaling points across the SS7 network regardless of whether they are directly connected. It includes such capabilities as node addressing, routing, alternate routing, and congestion control.

Collectively, MTP levels 2 and 3 are referred to as the message transfer part (MTP).

Signaling Connection Control Part

The signaling connection control part (SCCP) provides two major functions that are lacking in the MTP. The first of these is the capability to address applications within a signaling point. The MTP can only receive and deliver messages from a node as a whole; it does not deal with software applications within a node. TOP

While MTP network-management messages and basic call-setup messages are addressed to a node as a whole, other messages are used by separate applications (referred to as subsystems) within a node. Examples of subsystems are 800 call processing, calling-card processing, advanced intelligent network (AIN), and custom local-area signaling services (CLASS) services (e.g., repeat dialing and call return). The SCCP allows these subsystems to be addressed explicitly.

Global Title Translation

The second function provided by the SCCP is the ability to perform incremental routing using a capability called global title translation (GTT). GTT frees originating signaling points from the burden of having to know every potential destination to which they might have to route a message. A switch can originate a query, for example, and address it to an STP along with a request for GTT. The receiving STP can then examine a portion of the message, make a determination as to where the message should be routed, and then route it.

For example, calling-card queries (used to verify that a call can be properly billed to a calling card) must be routed to an SCP designated by the company that issued the calling card. Rather than maintaining a nationwide database of where such queries should be routed (based on the calling-card number), switches generate queries addressed to their local STPs, which, using GTT, select the correct destination to which the message should be routed. Note that there is no magic here; STPs must maintain a database that enables them to determine where a query should be routed. GTT effectively centralizes the problem and places it in a node (the STP) that has been designed to perform this function. TOP

In performing GTT, an STP does not need to know the exact final destination of a message. It can, instead, perform intermediate GTT, in which it uses its tables to find another STP further along the route to the destination. That STP, in turn, can perform final GTT, routing the message to its actual destination.

Intermediate GTT minimizes the need for STPs to maintain extensive information about nodes that are far removed from them. GTT also is used at the STP to share load among mated SCPs in both normal and failure scenarios. In these instances, when messages arrive at an STP for final GTT and routing to a database, the STP can select from among available redundant SCPs. It can select an SCP on either a priority basis (referred to as primary backup) or so as to equalize the load across all available SCPs (referred to as load sharing).

ISDN User Part (ISUP)

ISUP user part defines the messages and protocol used in the establishment and tear down of voice and data calls over the public switched network (PSN), and to manage the trunk network on which they rely. Despite its name, ISUP is used for both ISDN and non–ISDN calls. In the North American version of SS7, ISUP messages rely exclusively on MTP to transport messages between concerned nodes.

Transaction Capabilities Application Part (TCAP)

TCAP defines the messages and protocol used to communicate between applications (deployed as subsystems) in nodes. It is used for database services such as calling card, 800, and AIN as well as switch-to-switch services including repeat dialing and call return. Because TCAP messages must be delivered to individual applications within the nodes they address, they use the SCCP for transport.

Operations, Maintenance, and Administration Part (OMAP) TOP

OMAP defines messages and protocol designed to assist administrators of the SS7 network. To date, the most fully developed and deployed of these capabilities are procedures for validating network routing tables and for diagnosing link troubles. OMAP includes messages that use both the MTP and SCCP for routing. What Goes Over the Signaling Link

Signaling information is passed over the signaling link in messages, which are called signal units (SUs).

Three types of SUs are defined in the SS7 protocol.

SUs are transmitted continuously in both directions on any link that is in service. A signaling point that does not have MSUs or LSSUs to send will send FISUs over the link. The FISUs perform the function suggested by their name; they fill up the signaling link until there is a need to send purposeful signaling. They also facilitate link transmission monitoring and the acknowledgment of other SUs.

All transmission on the signaling link is broken up into 8-bit bytes, referred to as octets. SUs on a link are delimited by a unique 8-bit pattern known as a flag. The flag is defined as the 8-bit pattern 01111110. Because of the possibility that data within an SU would contain this pattern, bit manipulation techniques are used to ensure that the pattern does not occur within the message as it is transmitted over the link. (The SU is reconstructed once it has been taken off the link, and any bit manipulation is reversed.) Thus, any occurrence of the flag on the link indicates the end of one SU and the beginning of another. While in theory two flags could be placed between SUs (one to mark the end of the current message and one to mark the start of the next message), in practice a single flag is used for both purposes. Addressing in the SS7 Network TOP

Every network must have an addressing scheme, and the SS7 network is no different. Network addresses are required so that a node can exchange signaling nodes to which it does not have a physical signaling link. In SS7, addresses are assigned using a three-level hierarchy. Individual signaling points are identified as belonging to a cluster of signaling points. Within that cluster, each signaling point is assigned a member number. Similarly, a cluster is defined as being part of a network. Any node in the American SS7 network can be addressed by a three-level number defined by its network, cluster, and member numbers. Each of these numbers is an 8-bit number and can assume values from 0 to 255. This three-level address is known as the point code of the signaling point. A point code uniquely identifies a signaling point within the American SS7 network and is used whenever it is necessary to address that signaling point.

Network numbers are assigned on a nationwide basis by a neutral party. Regional Bell operating companies (RBOCs), major independent telephone companies, and interexchange carriers (IXCs) already have network numbers assigned. Because network numbers are a relatively scarce resource, companies' networks are expected to meet certain size requirements in order to be assigned a network number. Smaller networks can be assigned one or more cluster numbers within network numbers 1, 2, 3, and 4. The smallest networks are assigned point codes within network number 5. The cluster to which they are assigned is determined by the state in which they are located. The network number 0 is not available for assignment and network number 255 is reserved for future use. Signal Unit Structure TOP

SUs of each type follow a format unique to that type. A high-level view of those formats is shown in Figure 8.

Figure 8. Signaling Unit Formats

All three SU types have a set of common fields that are used by MTP Level 2. They are as follows:

Flag

Flags delimit SUs. A flag marks the end of one SU and the start of the next.

Checksum TOP

The checksum is an 8-bit sum intended to verify that the SU has passed across the link error-free. The checksum is calculated from the transmitted message by the transmitting signaling point and inserted in the message. On receipt, it is recalculated by the receiving signaling point. If the calculated result differs from the received checksum, the received SU has been corrupted. A retransmission is requested.

Length Indicator

The length indicator indicates the number of octets between itself and the checksum. It serves both as a check on the integrity of the SU and as a means of discriminating between different types of SUs at level 2. As can be inferred from Figure 8, FISUs have a length indicator of 0; LSSUs have a length indicator of 1 or 2 (currently all LSSUs have a length indicator of 1), and MSUs have a length-indicator greater than 2. According to the protocol, only 6 of the 8 bits in the length indicator field are actually used to store this length; thus the largest value that can be accommodated in the length indicator is 63. For MSUs with more than 63 octets following the length indicator, the value of 63 is used.

BSN/BIB FSN/FIB

These octets hold the backwards sequence number (BSN), the backwards indicator bit (BIB), the forward sequence number (FSN), and the forward indicator bit (FIB). These fields are used to confirm receipt of SUs and to ensure that they are received in the order in which they were transmitted. They also are used to provide flow control. MSUs and LSSUs, when transmitted, are assigned a sequence number that is placed in the forward sequence number field of the outgoing SU. This SU is stored by the transmitting signaling point until it is acknowledged by the receiving signaling point.

Because the seven bits allocated to the forward sequence number can store 128 distinct values, it follows that a signaling point is restricted to sending 128 unacknowledged SUs before it must await an acknowledgment. By acknowledging an SU, the receiving node frees that SU's sequence number at the transmitting node, making it available for a new outgoing SU. Signaling points acknowledge receipt of SUs by placing the sequence number of the last correctly received and in-sequence SU in the backwards sequence number of every SU they transmit. In that way, they acknowledge all previously received SUs as well. The forward and backwards indicator bits are used to indicate sequencing or data-corruption errors and to request retransmission. What are the Functions of the Different Signaling Units? TOP

FISUs themselves have no information payload. Their purpose is to occupy the link at those times when there are no LSSUs or MSUs to send. Because they undergo error checking, FISUs facilitate the constant monitoring of link quality in the absence of signaling traffic. FISUs also can be used to acknowledge the receipt of messages using the backwards sequence number and backwards indicator bit.

LSSUs are used to communicate information about the signaling link between the nodes on either end of the link. This information is contained in the status field of the SU (see Figure 8). Because the two ends of a link are controlled by independent processors, there is a need to provide a means for them to communicate. LSSUs provide the means for performing this function. LSSUs are used primarily to signal the initiation of link alignment, the quality of received signaling traffic, and the status of the processors at either end of the link. Because they are sent only between the signaling points at either end of the link, LSSUs do not require any addressing information. TOP

MSUs are the workhorses of the SS7 network. All signaling associated with call setup and tear down, database query and response, and SS7 network management takes place using MSUs. MSUs are the basic envelope within which all addressed signaling information is placed. As will be shown below, there are several different types of MSUs. All MSUs have certain fields in common. Other fields differ according to the type of message. The type of MSU is indicated in the service-information octet shown in Figure 8; the addressing and informational content of the MSU is contained in the signaling information field. Message Signal Unit Structure

The functionality of the message signal unit lies in the actual content of the service information octet and the signaling information field (see Figure 8).

The service information octet is an 8-bit field (as might be inferred from its name) that contains three types of information as follows:

Table 1. Common Signaling Indicator Values

TOP

The format of the contents of the signaling information field is determined by the service indicator. (Within user parts, there are further distinctions in message formats, but the service indicator provides the first piece of information necessary for routing or decoding the message.)

The first portion of the signaling information field is identical for all MSUs currently in use. It is referred to as the routing label. Simply stated, the routing label identifies the message originator, the intended destination of the message, and a field referred to as the signaling-link selection field which is used to distribute message traffic over the set of possible links and routes. The routing label consists of 7 octets that are outlined below in Table 2 (in order of transmission).

Table 2. Routing Label

Point codes consist of the three-part identifier (network number, cluster number, and member number), which uniquely identifies a signaling point.

Lines Versus Trunks TOP

The term line unfortunately refers to more than one type of circuit. In most cases, it includes a connection configured to support a normal voice calling load generated by one individual (typically about 10 minutes per hour for a business user). But in the case of a PBX , the term line usually corresponds to one connection from the PBX to a desktop. In the case of Centrex, a line is normally one physical connection from the customer site to the CO. With a key system, a line corresponds to one telephone number—but it might also be referred to as a trunk .

The term trunk normally refers to a circuit configured to support the calling loads generated by a group of users—possibly many thousands of users. Thus, a general-use circuit from a PBX to a CO would usually be described—and billed—as a trunk (but see DID lines). Connections between COs or higher in the network hierarchy would also be referred to as trunks. But note that these trunks are (or at least can be) physically identical to lines. Why then the different terminology? The ability of any given switching system such as a CO or a PBX to establish connections is limited. For example, although a PBX might be able to support 200 connections or ports, it might only be able to actually provide 80 paths at any one time. In such a case, if 80 people connected to 80 other people (some of them possibly off site), that would account for 160 of the ports; if any of the remaining 40 telephones or ports attempted to be serviced, it would fail. That is, a user could pick up the telephone and not receive a dial tone. Some systems are configured so that no such failures can happen. If only 160 physical connections were made to the PBX, then it could provide simultaneous service to all of them.

Normally, a PBX's connections to the CO are configured so that a much higher utilization than 10 minutes per hour is achieved on those ports; a primary benefit of a PBX is the ability to buy fewer telco connections than one has telephones. The CO must be configured so that it can provide connection services to such trunks at this higher utilization rate, thus using more of the co's overall switching and connection capacity (COs are not normally configured as nonblocking switches). So, the telco will naturally bill a PBX trunk at a higher rate than a single business line—even though the PBX trunk might be physically identical to that single line.

As a final comparison of the definition of line versus trunk, then would be as follows:

A line is an end point from a central switching service, such as a CO or a PABX. The line is represented as the end point on the pair of wires regardless of where the intelligence resides. A line carries one single conversation at a time on the physical channel capacity. It is a billable location for the telephone companies.

A trunk on the other hand connects between two intelligent switching systems. The trunk might be a single circuit carrying a single call at a time, or it might be a bundled service that is multiplexed and carries multiple simultaneous conversations. The difference is that a trunk will be used for switching and routing decisions from the switching offices (CO or PABX). The trunk is continually used rather than occasionally used. It is a billable address that can have additional sub addressing capabilities behind it. In a telephone company world, it is the connection between and among other offices in the hierarchy. In the private user (customer) world, it might be a single connection to the intelligent PABX from the CO. These distinctions offer some variations in billing and utilization. With this in mind, here are some common configurations.

Loop Start

The standard analog telephone line in a residence or business is 'loop start.' A PBX supplies the same interface to analog

Loop Start Interface

Calling Procedures TOP

When the line is idle, the phone is on-hook and the hook switch is open. The capacitor (C) in series with the ringing detector does not pass d.c. current (from the battery). With no d.c. path through the phone, the loop current through the loop plant is near zero.

To initiate a call, the caller picks up the phone (goes off-hook), closing the hook switch and letting the speech network (hybrid) draw current from the battery. This current powers the phone.

The CO switch sees this loop current as a request for service and a notice to prepare to receive a dialed phone number. When ready, usually within 3 seconds, the switch gives dial tone and no more than 70 ms later is ready for the caller to dial.

During dialing and call processing, depending on the make of the switch, the voltage from the network may reverse polarity, change up or down, or disappear briefly. If the CPE opens the loop during dialing for more than 1 ms, that may be counted as a pulse; an open circuit of more than 100 ms may disconnect the call.

Audible call progress signals are usually given to the caller: dial tone first; then ringing tone to indicate the far end has not yet answered, busy signal, or re-order (fast busy). At the end of the call, the phone is hung up (goes on-hook) which stops the loop current. An open loop tells the switch the line is idle again and to clear the call.

On lines with extra features that are activated by a 'hook flash' like call transfer, call waiting, or 3-way calling, any on-hook period over 1.5 s is considered a disconnect. An open interval of 300 to 1000 ms is taken as a hook flash signal to invoke a service.

When the switch has a call for a phone, the switch alerts the called party by applying an a.c. voltage (up to 170 V a.c. RMS) to the local loop. Alternating current passes through the capacitor C to activate the bell or other ringer. Alternating current into a standard analog phone energizes electromagnets that shake a clapper which strikes mechanical bells. More recent phones and PBXs may recognize an a.c. ringing current or merely the presence of an a.c. voltage of the expected frequency and amplitude (above some minimum voltage, as low as 40 V RMS for a 20 Hz ringer).

When the call is answered, the hook switch closes the d.c. loop. Battery current trips a relay or electronic monitor in the switch, which disconnects the line from the source of ringing voltage (in under 200 ms) and connects it to the talk path.

Ground Start TOP

An examination of the loop-start process reveals a problem if it is used a two-way interface between two switches, as when a PBX or Central Office \Voice Frame Relay Access Device (CO\VFRAD) is connected to central office lines. After a loop-start line has been seized by me calling switch, there is a periodof up to 4 seconds before the calling switch applies ringing voltage for the first time. During those 4 seconds the called switchdoesn't know that the line is in use and could attempt to make a call on the same trunk. This condition, called 'glare,' prevents theline from being used for either call attempt. Glare is important when both sides are automatic; loop start can be used for 2-way calling if one side is operated by an attendant who can deal with the condition.

To avoid glare, the 'ground start' (GS) interface is used between switches on two-way trunks. It has a more positive process to seize the line: one of the leads is grounded; ringing may be optional. The CPE side grounds the Ring lead, the network side grounds the Tip lead. In this context, the CPE is the PBX facing the CO\VFRAD, which acts like network.

If you look back at E&M-IV signaling, you'll see a strong resemblance to the basic ground start circuit. Each side applies battery, through a current detector, to a lead that the other side may ground. Unlike the E&M interface, GS is not fully symmetrical. Each side of the interface (FXS or FXO) is most often supplied in a different module, especially when ringing voltage is to be supplied by the FXS.

In the ground-start idle state the CO\VFRAD battery (+ side) is grounded by contact G. Contacts B and S, the ones used to ground the tip and ring leads, are open.

The current detector on each side of the interface registers when the other side grounds the lead and draws current from the battery. This ground signal is applied at seizure, so the called switch knows immediately that this line is not available. Grounding a lead is a positive indication of a new call; ringing voltage is not absolutely essential.

Ground Start Interface Ground Start Interface TOP

The procedure for the customer to originate a call is to close contact S, grounding the ring lead. At the same time the hookswitch places the phone circuit on the loop (tip and ring leads). Contact S draws current through the ring lead only, from the battery in the network side (in the CO\VFRAD).

When the FXS CO\VFRAD port (like the CO switch) recognizes the request for service, and is ready to receive dialed digits, it closes contact B to supply loop current and may open contact A. It may supply dial tone. The caller then dials the called number.

If the circuit has what's called 'floating battery,' the more modern type, contact A is closed only during idle. This ground completes the circuit with the ring lead so contact S can draw current. 'Conventional battery' in the CO\CO\VFRAD or switch is grounded at all times: contact G is always closed. Only one switch will exist in any circuit Interface.

When the FXS CO\VFRAD port or switch originates the connection on the interface it is 'terminating a call' (because the call originated somewhere else). The alerting action is to close contact B, grounding the tip lead, and apply ringing voltage to the ring lead (on top of the battery voltage). Note that the 2 s on, 4 s off cadence of ringing may mean that the tip line is grounded for some seconds before a.c. ringing voltage appears.

The PBX may answer immediately, on sensing the grounded tip lead, or it may react only to ringing voltage. In either case, the PBX (FXO port) answers by going on-hook. The loop current trips the ringing generator off, cuts through the talk path, and opens the battery grounding contact A (if using a floating battery).

After recognizing a new call, the GS interface operates like loop-start until both sides go back on-hook. The disconnect process depends on which side goes first and whether the battery is floating or conventional.

If the terminal (PBX) disconnects, it opens the loop by going on-hook (breaking the contacts of the hookswitch). If it is a conventional battery on the other side of the interface (in the FXS CO\VFRAD or CO switch), the PBX waits until ground is removed from the tip lead, then a short guard time, before returning the line to idle.

If the switch or FXS CO\VFRAD disconnects first, it opens contact B, stopping the loop current. A floating battery circuit will also disconnect the negative battery side from the ring lead for about half a second. The current detector in the PBX waits 350 ms to confirm that the open loop is not just a transient, then may wait a short guard time or immediately return the port to idle.

Ground start has other useful features that weren't imagined when it was introduced in the 1920s: The ground is removed by each side when it returns the line to idle. Positive indication of a disconnection before each new call helps automatic call distributors deal with high volumes of calls into hunt groups. Disconnect supervision prevents certain types of call fraud. Each end can tell when the other end has hung up. In England ground is called earth which leads to earth calling as the name for this interface type. But they've already heard the Joke about Earth calling Flash Gordon. TOP

E&M Signaling

One of the most common interfaces for PBXs and switches is E&M. In addition to the voice path, there are two signaling circuits, one to send and one to receive. Depending on your source, the abbreviation stands for:

·Earth and Magneto, early terms for ground and battery.

·E(ar) and M(outh), indicating what the switch or PBX equipment does on each lead when talking on the multi-wire interface to the FRAD or transmission line ;

·connector block positions that were marked alphabetically, E and M, rather than numbered, used years ago to install this type interface. The positions were codified in a standard 'practice' or instruction set for telephone company craftsmen, thus preserving the name.

Note that E&M leads may consist of either 2 wires (plus ground lead) or 4 wires (two balanced circuits). However these are not the wires referred to when calling an E&M interface '2-wire' or '4-wire':

E&M interfaces are characterized as 2-wire or ^-wire by the nature of the voice path, not the signaling leads. That is, voice can be carried on either a single loop of one pair of wires or on a separate wire pair transmitting in each direction. E&M modules are available to support both 2-wire and 4-wire voice paths. There are defined five different types of E&M signaling circuits, written as Roman numerals I to V. The number of physical wires used for signaling depends on the Type.

As noted for Type IV E&M, below, two CO\VFRAD may be hooked together directly (back to back) if the E lead on one CO\VFRAD is connected to the M lead on the other (the voice path must be connected as well).

Types I and III need an auxiliary circuit between them, to provide battery as well as cross-connect the E's to M's. Types II, IV, and V are easily connected together if the sB, M, sG, and E of one side are wired to the sG, E, sB, and M leads of the other, in that order. Ignore the sG and sB on Type V

Off the shelf interface converters are available to connect an E&M with practically any other interface, including another E&M. Some types and mixed types may have to be kluged together by adding a battery, ground, relay, etc.

The signal on the E or M lead is any of several conditions, again depending on the Type of E&M interface. The conditions called ground, battery, open, and closed. Typically, only two of the conditions are possible on any specific lead, one for each possible state of that lead: idle or asserted.

Note that signaling and hook switches in the diagrams that follow are irked N.O. (normally open) or N.C. (normally closed). In this section, normal means on-hook or idle, the state in which phones spend most of their lives. When a device goes off hook. the N.O. contacts close; the N.C. contacts open. TOP

The voice Frame Relay Access Device (VFRAD) is always assumed in this section on E&M to be transmission equipment, like the channel bank, and not the voice switch. It is the switch that has a mouth and ear. Therefore the E lead carries a signal from the CO\VFRAD to the PBX, and the M lead allows the PBX to signal the CO\VFRAD.

In cases where there is another E&M interface at the far end of the transmission line, any E&M signaling system conveys the M signal on one end to the E lead on the other. In this way there is a full duplex signaling path between the two switches or PBXs.

In the traditional telco world of channel banks and T-1 lines, the two states of the M lead (idle or asserted) are the only signaling messages. They translate into the two signaling states of the A bit: 0 = idle, 1 = asserted.

When the E&M interface is carried on frame relay, the FRADs on each end that are part of the signaling path have the same lead designation (E or \I) as the attached PBX:

(E on PBX) <E on CO\VFRAD) <frame relay net] <M on CO\VFRAD) <M on PBX) (M on PBX> (M on CO\VFRAD> [frame relay net> (E on CO\VFRAD> (E on PBX)

E&M is used in local loops. But rather than run so many wires (up to 8 for one voice circuit) over any great distance, the phone company uses signaling conversion devices to convey the state of the M lead on each end to the E lead at the other end. A CO\VFRAD performs this function on a frame relay network.

On analog circuits, the state of the M lead is converted to a tone signal carried on the voice path. The tone doesn't interfere with callers because the presence of the tone indicates the line is idle: the tone must be absent if the circuit is in use ('seized' or 'off hook').

Often a type of E&M interface at one side is allowed to interact with some other interface, like loop start. These applications may use proprietary features. TOP

The E&M interface is defined from the switch to transmission equipment, with the assumption that there is another switch at the other side of the transmission trunk. For the PBX to seize the trunk, it asserts the M signal when the E lead is not asserted (E asserted means the trunk is busy, or seized from the other end). When the far switch responds, by asserting its M lead, the originating switch sees that response signal on its E lead.

If dialing information is to be transmitted between the switches, the called end may indicate its readiness to receive digits in one of two ways:

·'Wink start' ROM: The called end, when ready to receive dialing information, sends back a 'wink'. This is a signal of at least 140 milliseconds, asserted on the far-end M lead, that appears on the originating end's E lead. The Called Switch also may return audible dial tone on the voice path at the same time. The ANSI standard for PBXs (EIA-464) allows dial tone on the voice path to be substituted for the wink signal, if the other PBX will accept it; a configuration issue. The calling end then dials, either by toggling the M lead or by transmitting DTMF tones on the voice path.

·Delay dial trunk: when the called end sees the E lead active, it immediately (<300 ms) asserts its M lead (goes 'off hook') until ready to receive dialed digits. When ready, it returns to 'on hook' and may or may not supply dial tone. The calling end then dials, using pulse or tone. When the called switch is not busy and can respond quickly, both wink-•start and delay-dial will behave similarly and may be difficult to tell apart. A busy switch will be ready to receive dialing at a certain time, regardless of e trunk signaling type. That is, a 'Delay' pulse will be longer than a 'Wink' pulse but both will end at the same time.

A third trunk type doesn't offer acknowledgement to a seizure: —

·'Immediate start' trunk: the originating end starts dialing a short, \ed time after seizing the trunk. Dial tone from the called switch is optional. TOP

In all cases the called PBX will assert its M lead when the called extension answers. The calling PBX sees this on its E lead.

There is no hard rule against any of these three trunk types having an interface with any of the five types of E&M. However, due to the delay across a frame-based network, it would be difficult, perhaps impossible, for two switches to use delay-dial on any E&M interface. Delay-dial requires a quick response to each seizure, in a time that may be less than the round-trip propagation delay across the frame relay backbone.

Originally, the M lead was toggled between two states (asserted and not asserted) to transmit dial pulses (rotary dialing). By mapping pulse dialing into M lead states, and that into signaling bits, it is possible to transport pulse dialing under the FR Forum Implementation Agreement for VoFR, using the 'dialed digits' subframe payload type.

Signaling between E&M interfaces may involve fairly critical timing of events The length of a wink is closely controlled to distinguish it from a disconnect (hang-up) followed immediately by another call request. Replicating dial pulses may have to preserve the ratio of on to off, which varies by country. In the US, the break (open loop circuit, on-hook state) is held close to 61% of the pulse time; in other countries the make: break ratio can be closer to 1:1.

Timing for these older, analog interfaces is the reason the FRF VoFR implementation agreement includes an emulation of ABCD signaling bits. In moving to the future, with ISDN and digital interfaces, the signaling will be based on messages like those in Q.931/Q.933 and Q.SIG from the ITU.

Once the near end has seized a trunk (and been acknowledged by the far end if the trunk type is wink start or delay dial) the originating end will transmit dialing. While it is still possible to toggle the M lead, switches today most often use DTMF (dual tone multi-frequency), the push-button dialing signals. DTMF is much faster than rotary dialing. Between central office switches, there is also a telco system of multi-tone signaling that is not included in the first version of the VoFR IA.

After the called station (telephone) answers (goes 'off hook') the called switch asserts the M signal lead, which appears on the E lead at the calling end. Both E and M remain asserted while the call is in progress. When either switch detects an 'on hook' from its station, it drops its M signal (causing E to drop at the other end). The second switch, if its phone is not hung up, should give it dial tone and behave as if a call were originating. TOP

Note that asserting a signal lead might result in ending a signaling tone on the voice path, as happens with interoffice analog E&M signaling.

E&M Interface, Switch to Trunk Trunk Dialing

Type I: 2 E&M wires, ground

The PBX supplies battery voltage for both E and M in Type I signaling. Battery on the E lead is presented through a current detector. The positive side of the battery is grounded. The CO\VFRAD asserts the E signal by grounding the E lead at its end, drawing a current that is detected by the PBX. The transmission equipment (CO\VFRAD) terminates the M lead through a current detector tied to ground. The PBX asserts M by switching its end of the M lead from ground to battery, delivering a current that is detected in the CO\VFRAD. Type I is common in the US. It is convenient for transmission equip ment, like CO\VFRAD because the PBX supplies battery for both E and M The circuitry in the CO\VFRAD is passive.

E&M Type 1 Signaling

Type II: 4 E&M wires TOP

Because the PBX supplies battery on both leads to all trunks in Type I, the unbalanced ground current may become large and cause interference. Type II avoids this problem by making both E and M leads balanced. That is:

·The CO\VFRAD does not ground the E lead locally, but at the PBX through a separate lead, sG (signal ground).

·The battery source for the M lead is in the CO\VFRAD, delivered to the PBX on the sB lead (signal battery); the PBX closes the loop to assert M (there is no grounding of M at idle).

Type II is convenient because it is almost symmetrical—except for the location of the detector in the transmission interface being on the grounded leg, where it is on the battery leg in the switch. Still, two Type II devices of the same kind (two switches, or two CO\VFRAD may be connected together directly (locally) by crossing E to M and M to E (with their signal grounds). However, Type II also may be inconvenient because it is symmetrical. Now both sides have to supply battery. This is not a problem if there are two real switches connected by 6 or 8 copper wires (2- or 4-wire voice path). However, a VoFR device would need a suitable voltage source (at least -21 V), which is more than the 12 V commonly found in electronics equipment.

E&M Type ll Signaling

Type III: 4 E&M wires TOP

The E lead on Type III is unbalanced, the same as Type I. M signaling is the same as Type I (switch from ground to battery to assert) except that both are provided by the CO\VFRAD (via sG and sB leads).

E&M Type lll Signaling

Type IV: 4 E&M wires

The switch side of Type IV is identical to the Type II, but the transmission side has the detector in the battery leg of the M circuit. Both sides are identical.

E&M Type lV Signaling Type V: 2 E&M wires, ground

The E lead of Type V is the same as Type I. For the M lead, the CO\VFRAD supplies battery through a current detector. The PBX grounds M to assert a signal, drawing a current that is detected by the CO\VFRAD. The E lead is symmetrical with the M lead. Thus they may be crossed between similar equipment types to create a back-to-back connection. Type V is popular in Europe and the UK.

E&M Type V Signaling

Foreign Exchange (FX) Signaling TOP

Think of FX as an extension cord for the 2-wire analog interface between a switch and a telephone . When extending a line from a PBX, FX is also called an ' If this extension cord runs through a frame relay network, the voice interface functions must be supplied by CO\VFRAD at each end. The CO\VFRADmimic a phone at the switch and a switch at the phone. In other words:

1. the 'customer premises equipment interface' (CPE-I) at the customer end of the loop plant must be duplicated in the CO\VFRAD (FXO) on the CO/PBX side of the frame relay link;

2. the switch interface at the CO end of the loop plant must be duplicated in the CO\VFRAD (FXS) on the CPE side of the frame relay link.

At one end of the cord (the office end) the switch supplies battery, generates ringing voltage, and accepts dialing. The transmission equipment (the CO\VFRAD) at that end of the FX extension cord is 'FXO' for foreign exchange, office.' At the other end of the transmission system, the phone accepts battery current and ringing voltage, and generates dialing. The transmission equipment there (CO\VFRAD) presents an 'FXS' interface ('foreign exchange, subscriber' or 'station') to face the station equipment.

The name describes what the port faces, not what it is (office or subscriber). To send a call to a phone over the frame relay network, the PBX rings an extension port as if the phone were attached locally. The local CO\VFRAD, emulating the phone, supports the FXO functions: it absorbs the ringing current, recognizes it as ringing, and sends a call request or start ringing message to the remote CO\VFRAD TOP

When a call request message arrives from the FR network, the remote CO\VFRAD applies ringing voltage to the 2-wire voice interface. When the called phone goes 'off hook' the CO\VFRAD halts ringing and delivers loop current, returning an 'off hook' signal to the calling end. The voice path is established between the CO\VFRAD

When the far end's 'off hook' message arrives at the originating (calling) CO\VFRAD, it goes 'off hook' itself. This action draws loop current from the PBX and stops ringing. The PBX then opens its talk path to the local CO\VFRAD, which has already set up a talk path to the remote CO\VFRAD. Now, audio energy received by the remote CO\VFRAD on the analog interface is treated as voice: digitized, compressed, packetized, encapsulated in frame relay, and transmitted via the FR network. At the local end the process is reversed to present analog speech to the PBX.

To originate a call at a remote extension, that phone goes off-hook, which draws loop current to tell the remote CO\VFRAD someone wants to place a call. The remote CO\VFRAD signals the local CO\VFRAD (at the switch) via a signaling frame, which uses the same channel ID as the voice path.

The local CO\VFRAD goes 'off hook' to draw loop current from the PBX. This causes the PBX to send dial tone, which is signaled to the remote CO\VFRAD. DTMF or pulse dialing from the remote station arrives as coded messages which are translated by the local CO\VFRAD into re-created DTMF tones or dial pulses. When the remote extension returns to *on hook' the remote CO\VFRAD signals the local CO\VFRAD to go 'on hook' also, opening the loop to stop drawing current. This indicates to the PBX an 'idle' extension. The call is cleared.

A straight FX function is 2-wire analog at both ends, FXS to FXO. With signaling conversion in the CO\VFRAD an FX interface (FXS or FXO) may be mixed with an E&M or digital interface at the other end.

On some equipment the voice path is 4-wire. The two paths are transformer coupled so that each pair may be used as a signaling lead. The call processing procedures are the same as for a single loop, as described below, substituting the A and B leads for tip and ring. With a 4-wire interface, a signal in each direction feeds directly into an amplifier where necessary. A hybrid is not needed to split the signal—eliminating the prime source of echo. Lack of echo is one of the advantages of this interface type, particularly when the network is 4-wire also.

FXS/FXO Extends Analog Interface 4-Wire Interface for Loop Procedure

DID TOP

DID refers to direct inward dialing. From a caller's point of view, this service is in place if the caller can dial a 10-digit number from the outside, and reach a specific individual without operator (live or automated) intervention. Thus, Centrex normally inherently supports this capability without any additional configuration—everyone already has their own telephone number. A true key system (where telephone numbers are normally shared) can only do it if any given telephone number has only a single appearance.

But DID is usually referred to in the context of a PBX. It is a specific PBX feature that must be enabled and configured, with elements set up both within the PBX and also with the telco. Consider as an example a new site intended to support 1100 employees, each with his or her own telephone connected to a PBX.

The first step in arranging DID is to reserve the telephone numbers for all those employees. Let's say that the main company telephone number is 555-1234. The Telecommunications manager will request a block of DID numbers from the telco, probably about 2000. The telco might say, Your DID numbers are 555-2200 through 555-4199. Notice that while there is a good chance the block will have the same exchange as the main number, it probably will not include (and one would not want it to include) the main number. The company will pay for these numbers on a monthly basis, but nowhere near as much as would be required for actual telephone lines. So far, the only thing arranged is the reservation of the block of numbers themselves. These numbers will not be given out by the telco to anyone else. The Telecommunications manager will assign each employee one of the numbers in the DID block.

Next, the Telecommunications manager must determine how many trunks (or DID lines) in the trunk group will be required to support the calls from outside to the company's employees. These are inbound only, and are in addition to the normal in-out or inbound trunks that serve the main operator, so they must be engineered to a very low level blocking indeed. With DID, the telco passes on to the customer PBX the responsibility of handling answer supervision (e.g., busy signals).

So, if an external customer calls Jane at extension 2313, the customer will dial 555-2313. The telco CO will seize the next available trunk in the DID group (if no trunk is available, the caller will receive a busy signal), and signal along it that there is a call for extension 2313. At that point, if extension 2313 is busy, the PBX must deal with it; the CO is merely passing along the signals. Possible PBX actions include forwarding to a message center, generating a busy signal, or forwarding the call to a specified alternate extension.

DID is most often used to reduce or eliminate the manpower required for a central answering position. The more calls that customers can place directly, the fewer must be answered by the company operator. On the other hand, some companies prefer to have all incoming calls answered by someone trained in how that company wants its telephones to be answered (e.g., Thank you for calling Kay's deli! How can I help you vs. Hello?). It would generally be a mistake for a Telecommunications manager to make decisions regarding whether DID is to be implemented without consulting with company management. TOP

DOD

DOD refers to direct outward dialing. If an employee can dial and reach an outside number without internal operator intervention, then the company has implemented DOD. In the past, when less-sophisticated telephone systems were available, it was not uncommon for a company to route all of its outbound calls through an internal operator. The operator's responsibility was both to screen calls (no, you may not call Australia from that telephone) and to route the calls over the appropriate facilities (e.g., the right WATS line—see below). With the advent of modern PBXs and Centrex, such limitations can be programmed, if desired, on a telephone by telephone, or even user-by-user basis, eliminating the requirement to involve an operator in outbound calls. DOD is a term not often used these days because few companies consider not providing it. TOP

FX

FX, not to be confused with FAX, refers to a foreign exchange circuit. In this case, foreign refers to a CO other than one's own local CO, not to a location outside the country.

Consider the case of an airline that wishes to locate all of its reservations clerks in Atlanta. It cannot expect all of its customers to pay long-distance charges to make reservations. What are its alternatives? One possibility is a group of 800 circuits. Indeed, it will probably have a large number of those, but 800 trunks cover large areas (and are priced accordingly). What about service for customers calling from large, high-density metropolitan centers, such as Chicago? Perhaps a more focused service might be more cost-effective.

Think of an FX line (or trunk) as two-thirds of a dedicated point-to-point (or tie) connection. It starts at the customer's location, connects to the local CO, and extends from there to another foreign CO anywhere in the country. There is a fixed monthly charge for all that mileage; but there are no usage-sensitive charges for these miles. At the foreign CO, it is open. It has a telephone number associated with that foreign CO. Calls made to that number ring at the customer's location. Calls made from the customer's location over the FX line emanate from the foreign CO, incurring only local charges for the call from the foreign CO to the called location.

FX lines are often used by companies to provide a local number that customers can call in cities where those companies do not in fact have offices. In the airline's case, it could arrange a group of FX lines from its Atlanta offices to a Chicago CO. All of the lines could share one Chicago local telephone number. People from anywhere could call the number, but normally only Chicagoans would, because it would appear only in their telephone book—and it would be a local call only for them. If the airline wished to allow it, service representatives could also place calls from Atlanta to Chicago over the FX lines. The calls would be billed as though they were placed from within Chicago. Perhaps calls notifying customers of changed flight information might be placed this way. TOP

OPX

OPX refers to off premises extension. An OPX line permits a telephone not at a company's location to function to all intents and purposes as though it is located at the company's location. This capability becomes particularly interesting with the recent increase in telecommuting. Suppose an employee plans to work at home. One of the problems to overcome in such a case is the isolation such a worker might experience. Providing the employee a telephone that looks like an internal line at the company might help to reduce the problem. Others calling the line within the company will dial an internal extension, which will ring at the employee's home; if the employee wishes to make a long-distance call, he or she usually just dials y and the rest of number just as though they were at a desk at the company's location.

As with an FX line, an OPX connects from the company's location to the local CO, then continues via whatever intervening COs are necessary until it terminates directly on a telephone at another location. A key difference from an FX, however, is that on the PBX an OPX is connected and configured as a telephone rather than a trunk. This results in a limitation on the type of service provided: normally, only an analog telephone can be used at the end of an OPX because the digital signaling between a PBX and its proprietary telephones will probably not successfully make it through the various analog and digital circuits that make up the OPX. This limitation is not normally a show-stopper; rather, it just imposes on the telecommunications manager the need to configure the PBX to support a certain number of analog telephones as well as the digital telephones that might be used in-house. TOP

WATS

WATS is an abbreviation for wide area telephone service. WATS lines come in two flavors: in-WATS and out-WATS. Another name for in-WATS is 800 service. When most people refer to a WATS line, they mean an out-WATS facility. Both services are merely billing arrangements for reduced billing of long-distance calls based on a fixed monthly fee and discounts for larger calling volumes. 800 service also has the characteristic of reversing the charges to the called party.

Historically, WATS lines have been separate facilities (physically identical to local PBX trunks or private lines). Their geographic coverage was also banded; thus, one might have had a WATS line that only reached adjacent states (band 1), or all of the lower 48 United States (band 5), or some intermediate variation. For out-WATS, either one's PBX had to be smart enough to recognize the dialed area and choose the correct outgoing facility, or users had to dial special codes to select the right WATS line. Any given band included all closer bands (but, of course, billed the calls at the higher rate for the wider band), and this caused a certain amount of difficulty in either configuration or training. Because different physical facilities went to different regions, this also resulted in a traffic engineering nightmare. This complication was all the more unreasonable because WATS calls (both in and out) are handled identically to all non-WATS calls; WATS is really only a bulk billing arrangement for calls that would otherwise be considered direct distance dialing (DDD or toll calls).

WATS service has never been free, although some of the older tariffs did have points where all calls above a certain (rather large) volume were free. Those tariffs are long gone; all calls now cost on a per-minute basis. The only variable is the per-minute charge, which does decrease as the calling volume increases. One significant improvement is that WATS-type volume discount billing can now be set up on existing trunks; no longer is it required to have separate facilities into the local CO to have such an arrangement.

TOP

Home