An Aspect of the Dialing Behavior of Subscribers and Its Effect 

 on the Trunk Plant 



By CHARLES CLOS 



Introduction 



Tn\URING the war it became necessary for the Bell System Companies 

 ^^ to lower many service standards. Among these was the standard for 

 the provision of trunks for handling subscriber-dialed calls. In the interest 

 of economy the number of trunks for a given volume of traffic was lowered. 

 It is evident that for any given case there is a lower limit to the number 

 of trunks that should be provided for handling subscriber-dialed calls. 

 Below this limit congestion of calls gets beyond control. The control of 

 congestion is important. In the case of operator-handled calls it is possible 

 to control congestion by filing tickets and placing calls in an orderly fashion. 

 In the case of subscriber-dialed calls the subscriber may with impunity 

 make many, indeed very many, successive dialing attempts to complete a 

 call that is blocked due to a shortage of trunks. If, in a particular office 

 enough subscribers do this simultaneously, a sender shortage may develop 

 with its resulting reaction on the whole office, 



From the foregoing it is evident that the standard of service for providing 

 trunks in trunk groups handling subscriber-dialed calls is of importance. 

 During the war years, the New York Telephone Company undertook a 

 study to determine the limits below which it would be undesirable to degrade 

 the service. This study was designed to test the reasonableness of the 

 reduction in the inter-office trunk standard from the pre-war basis of pro- 

 viding enough trunks to delay only one out of a hundred calls in the busy 

 hour to a wartime basis of providing enough trunks to delay two calls in 

 every hundred during the busy hour. The conclusion from this study was 

 that it was safe to use wartime standards. 



The study reported herein is an analysis of the effect of repeated attempts 

 when subscriber-dialed calls are blocked due to trunk shortages. The data 

 upon which the results are based indicate that dial subscribers after en- 

 countering a busy condition make new attempts sooner and much more 

 often than has been generally believed. The results indicate that one can 

 reconstruct what happens when trunk groups carrying subscriber-dialed 

 calls encounter serious overloads and that trunk capacity tables for such 

 situations can be developed. 



The study is based on extensive service observations taken at the New 



424 



