THEORIES FOR TOLL TRAFFIC ENGINEERING IN THE U. S. A. 423 



and economical service between customers is a problem of great magni- 

 tude; it has engaged our planning engineers for many years. 



There are now 52 million telephones in the United States, over 80 per 

 cent of which are equipped with dials. Until quite recently most telephone 

 users were limited in their direct dialing to the local or immediately sur- 

 rounding areas and long distance operators were obliged to build up a 

 circuit with the aid of a "through" operator at each switching point. 



Both speed and economy dictated the automatic build-up of long toll 

 circuits without the intervention of more than the originating toll oper- 

 ator. The development of the No. 4-type toll crossbar switching system 

 with its ability to accept, translate, and pass on the necessary digits (or 

 lujuivalent information) to the distant office made this method of oper- 

 ation possible and feasible. It was introduced during World War II, and 

 now by means of it and allied equipment, 55 per cent of all long distance 

 calls (over 25 miles) are completed by the originating operator. 



As more elaborate switching and charge-recording arrangements were 

 developed, particularly in metropolitan areas, the distances which cus- 

 tomers themselves might dial measurably increased. This expansion of 

 the local dialing area was found to be both economical and pleasing to 

 the users. It was then not too great an effort to visualize customers 

 dialing to all other telephones in the United States and neighboring 

 countries, and perhaps ultimately across the sea. 



The physical accomplishment of nationwide direct distance dialing 

 which is now gradually being introduced has involved, as may well be 

 imagined, an immense amount of advance study and fundamental plan- 

 ning. Adequate transmission and signalling with up to eight intertoll 

 trunks in tandem, a nationwide uniform numbering plan simple enough 

 to be used accurately and easily by the ordinary telephone caller, pro- 

 ^ ision for automatic recording of who called whom and how long he 

 talked, with subsequent automatic message accounting, are a few of 

 man}^ problems which have required solution. How they are being met is 

 a romantic story beyond the scope of the present paper. The references 

 given in the bibliography at the end contain much of the history as well 

 as the plans for the future. • 



2. PRESENT TOLL TRAFFIC ENGINEERING PRACTICE 



There are today approximately 116,000 intertoll trunks (over 25 miles 

 in length) in the Bell System, apportioned among some 13,000 trunk 

 groups. A small segment of the 2,600 toll centers which they interconnect 

 is shown in Fig. 1. Most of these intertoll groups are presently traffic 

 engineered to operate according to one of several so-called T-schedules: 

 T-8, T-15, T-30, T-60, or T-120. The number following T (T for Toll) is 



