Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
SECTION 4 NETWORK-CONTROL SIGNALING INTRODUCTION The network-control signaling functions are associated with the initiation, placing, answering, and charging of calls over the switched network. Malfunctions can causeincompleted calls, or calls completed to other than the intended terminaL Processing such calls reduces the capacity of the network to serve "normal" calls. The effects of these malfunctions may be felt by all users of the system, not just those originating and answering imperfect calls. The present state of the switched telephone network does not permit easy identification of the source of this kind of malfunction; that is, to locate it as occuring in the subscriber's station or in the central office. Carrier-maintenance personnel, tests, and administrative procedures become involved in the attempt to localize these malfunctions as they come to light. . Consequences of Improper NetworkcControl Signaling The consequences of improper network signaling pervade the entire network and can be grouped into the following categories: (a) Wasteful use of central office and transmission facilities (b) Annoyance to other users (c) Incorrect billing (d) Wasted testing and maintenance effort (e) Added administrative expense Following are examples of each category: 1. Wasteful Use of Central Office and Transmission'Facilities. Wrong numbers caused by a faulty network-control signaling unit represent a waste of switching equipment and a source of annoyance to those who are wrongly called. - 33 -
- 34 - Furthermore, a wrong number resulting from faulty signaling can cause a call to end up in the wrong city. In the near future, a wrong number may tie up, for a time, a trans- atlantic cable or satellite trunk connection. There are other sources of faulty control signaling. If, when a call is completed, the switch hook contacts fail to open properly, or some extraneous impedance remains bridged across the line, it is arranged that the connection will release after a time-out of thirty seconds. This is thirty seconds during which the circuits are not available to other users. 2. Annoyance to Other Users. In the example mentioned above, in which the call is not released properly, the user himself will be unable to place calls during this interval and others trying to reach him will receive busy signals. 3. Incorrect Billing. On a two-party line, the billing equipment at the central office recognizes which party is making a call because there is a high-impedance DC connection to ground on one side 9f the line. If the connection is not made in the telephone, or if the telephone is installed or maintained improperly, the wrong party will be charged for some calls. On lines with more than two parties, more complex party identification schemes are used, which depend upon the telephone instrument having particular identifying characteristics that differ from the instruments on the same line. 4. Added Testing and Maintenance Effort. When excessive wrong numbers occur, action must be taken to identify the source. It might be on the loops, in the line circuit, or in the central office. On the other hand, it might be in the network-control signaling unit. The user unable himself to determine where the problem is located will normally call the telephone company. Faulty network contact signaling often shows up as an intermittent trouble. These are the hardest to trace and to diagnose. ·5. Added Administrative Effort. Improper network-control signaling can result in customer demands for credit against his telephone bill due to false charges. Since the source of the trouble, as previously mentioned, is difficult to trace and correct, the added administra- tive effort required can be considerable. Conclusions Improper network-control signaling leads to inaccurate billing, wasteful use of the telephone plant and administrative effort, as well as
- 35 - In planning for the use of user-owned network- a nnoyance tn other users. the quality of network-control signaling control signaling devices, must be preserved. AT&T Company·Expetiance with NetwotkCC6ntt6lSignaling The only available reliable source of information to the Panel on network-control signaling·is experience With this function in the operation of the sWitched telephone network. In this section, information and data furnished by AT&T are summarized. Dial-Pulse Signaling Network-control signaling failures are largely related to the familiar rotary dial. Sources of trouble here are: Finger wheel and stop (a) (b) Contact Mechanism (c) (d) Noise (e) Other The dial mechanism itself was the most frequent source of difficulty. The mechanism is required to operate at speeds nominally between 9.5 and 10.5 pulses per second and with a percentage break between 58 and 64 percent. Generally, the units used by the Bell System fail in such a manner as to fall outside the percentage-break tolerances. This type of failure can lead to dialing wrong numbers. Data on units supplied by others is sketchy. AT&T and Bell Laboratories, however, had reported experience with some equipment they have tested and found deficient. For example, one unit tested had a low-priced "antique" telephone with these two faults: (a) Low ringer impedance (b) Percentage break 67 percent outside allowable range of 58 - 64 percent The first fault is attributable to poor design. The second may indicate either poor design or maladjustment. Bell has also tested commercial answering machines and repertory dialers. Some answering machines had the characteristic of failing to disconnect promptly. One
- 36 - repertory dialer tested exhibited· improper percentage break as a function of line voltage, missed digits on low line voltage, and had inadequate interdigital time. A second repertory dialer exhibited dial speed and percentage-break characteristics that aged beyond specified limits. On the other hand, general experience with· telephones made by reputable manufacturers of telephonic equipment has indicated that the quality of network-control signaling units is on a par with those supplied by Bell. No comparative statistics are available. Based upon the statistics provided AT&T, the mean ·time between failures for Bell station sets is 8.5 years. The mean time between failures for rotary dials is 46 years and for ringers 59 years. The combination of rotary dial. and ringer has a mean time between failures of 26 years. It is this kind of performance, or better, that must be realized where new devices and systems are attached to the telephone system if present network-control performance levels are not to be degraded. Touch-Tone Signaling Touch-tone signaling uses two tones per digit generated by pushing buttons on the telephone. One tone is selected from four fre- quencies between 697 and 941 Hz. The second tone is selected from four frequencies between 1,209 and 1,633 Hz. Both tones must be received by the central office for it to be accepted as a valid digit. Frequencies have a + 1.5 percent tolerance. Output power is made a function of line current-to regulate the received power at the central office for various loop lengths. Other tolerances are specified to hold the two sets of tones at appropriate power levels. The unit must operate within tolerance over a -30°C to +55°C temperature range and during its service life. Reliable statistics on types and frequency of failures are not available on touch-tone dialers. Failure of the multi-frequency dialer due to improper frequency or power level, for example, will not be interpreted .by the central office as a wrong number. The more likely condition is a register time-out due to its failure to recognize all the transmitted digits. This use of central office facilities is considered relatively insignificant as a harmful effect when compared to harmful effects due to malfunctioning rotary dials. We conclude failure of touch-tone (multifrequency) signaling to be considerably less harmful to the network than failure of dial-pulse signaling.
- 37 - Haintenance Data In the switched telephone network, network-control signaling is exercised by the customer through the telephone instrument and over his wire loop to the central office. It is pertinent, therefore, to examine available data on station troubles and the costs associated with mainten- ance and trouble clearing. The following data were supplied to the Panel by representatives of AT&T. In 1967, Bell had 42,586,551 customer-trouble reports - 27,392,760 troubles were found as a result. These troubles broke down as follows: 8,608,962 30.8% Station set Other station equipment 4,302,696 15.4% Station wiring 4,802,760 17.2% Outside plant 5,390,924 19.3% Central office 2,485,913 8.9% Customer action 6.4% 1,801"505 System study of station troubles made in 1966 showed the ~~ell folloWing breakdown: Trouble rate/100 stations/month 0.21 Cord 0.18 Dial 0.14 Ringer 0.12 Key and lamp Mounting and plastic 0.12 0.08 Circuit 0.03 Receiver Transmitter 0.03 0.03 Other Whether these data reflecting carrier experience would be valid for customer-furnished station equipment, would depend on the performance of this equipment relative to that furnished by the common carrier. It would also depend on the extent of use of touch-tone control instead of rotary dial. FAULTY NETWORK-CONTROL SIGNALING WITH USER-OWNED EQUIPMENT It is difficult to evaluate the effect of interconnection on network-control signaling, since it is not known at present what precrse instrumentalities users will employ for this function. Network-control signaling performance is closely related to the very detailed design and performance of the device used (switch hook, rotary dial, touch-tone pad).
- 38 - The best that can be done, therefore, is to cite present experience of the carriers using their own devices. Starting from this as a reference point, it may be postulated that devices owned and used by customers will be either (a) as good as, or (b) poorer than, these carrier-furnished devices. The consequences of these assumptions are drawn in the following section. ECONOMIC PENALTIES OF NET CONTROL SIGNAL-DEVICE FAILURES Data on Bell System rotary dial and ringer units show a mean time between failures O1TBF) of 26 years. This is equivalent to a failure rate of 0.0385 per year. Except for the special case in which competent maintenance personnel are continually at hand, trouble visits will be required and costs will be incurred and must be paid for. Some vendors and users might be satisfied with a seemingly reasonable, though lower, MTBF. Reliability, however, has a profound impact on network operation and cost. Based on a large volume and using the Bell System experience of $15 per maintenance visit, Table 1 shows the annual average per phone cost for maintenance alone as a function of MTBF. The distribution of this cost between the user and the carrier cannot be determined at this time; however, it represents a substantial factor to be considered in specifying the performance of network-control signaling units. TABLE 1 Annual Maintenance Cost MTBF Allocated·to Each Phone 26 years .57 $ .75 20 years 15 years 1.00 10 years 1.50 5 years 3.00 1 year 15.00 Another cost (to the carrier) associated with improper network- control signaling failures is that attributed to wrong numbers, wrong toll charges, etc. It is difficult to estimate the frequency of such occurrences as a function of MTBF. A third cost associated with network-control signal-unit failures
- 39 - is that due to false calls for assistance by the user. Where limited free interconnection has been permitted in the past, it has been the experience of the carrier that he is frequently called to perform the maintenance when, in fact, the interconnected equipment is at fault. This phenomenon can be expected to persist with any form of interconnection in which a specific ~nterface between vendor equipment and the telephone company is not clearly defined. The three types of costs described are a function of the MTBF of the net-control signal unit. The costs are very significant when evaluated in terms of a large number of subscribers. These costs will be borne by both users and the carrier; since some costs cannot be easily allocated. CONCLUSIONS Net-control signaling is a critical element, and a high order of reliability is necessary to avoid loss of net performance and excessive costs to both carrier and user~
