310 



THE BELL SYSTEM TECHNICAL JOURNAL, MARCH 1953 



Table I 



Group Classification 



Heavy demand individual line subscribers (such as 



doctor and professional services 



Medium demand individual line subscribers (such as 

 small business and some residential subscribers) . . . 



Light demand individual line subscribers (mainly resi- 

 dential subscribers) 



P.B.X. line subscribers (such as large businesses, 

 hotels, railroads, etc.) 



Average'^ Calling Rate Per 

 SubscriberjDuring Busy Hour 



5.0 

 0.85 

 0.02 

 3.0 



before dial tone, and forgetting the number. Such actions produce 

 waste usage of equipment within the switching system, and their studj^ 

 is pertinent to producing throwdown data. To simpUfy this study, all 

 subscriber actions involving the alerting of central office equipment 

 are designated "subscriber starts" and divided into four categories as 

 given in Table II. 



Since "no dials" and "partial" dials are largely due to subscriber errors 

 in properly originating calls, many of these calls wdll be originated upon 

 discovery of the error. False starts, on the other hand, are attributed to 

 accidental origination with no intent to place a call. A flow chart illus- 

 trating these actions is show^n in Fig. 6. The importance of this sub- 

 scriber behavior is indicated by the percentage of waste usage calls (FS, 

 ND, and PD) to ultimate good calls. Pen recorder tapes taken at par- 

 ticular central offices showed waste usage calls at 30% and good calls at 

 80.5% of ultimate good calls. For these specific cases false starts repre- 

 sented 7.5%, partial dials 7.5%, and no dials 15% of ultimate good calls. 



Table II 



Category 



Description 



Good calls 

 False starts 

 No dials 



Partial dials 



Calls on which the subscribei; waits for dial tone and then 

 dials the required full code 



Calls, on which a sender or register is seized, but which are 

 abandoned in less than two seconds without dialing 



Calls lasting longer than false starts but on which no dialing 

 occurs. These calls may exceed a certain length "time- 

 out" period and be given a distinctive tone 



Calls on which less than the required full code is dialed. 

 These calls may be held beyond a certain length "time- 

 out" period and be given a tone 



