444 



BELL SYSTEM TECHNICAL JOURNAL 



The effect of using numerals in this way is that 90% of the calls encountering 

 all trunks busies appear as subsequent attempts. 



The numeral in the first place of each of the third set of random numbers 

 was used to establish the time interval for resubmitting each call. The time 

 intervals were developed from the data on Fig. 4 by dividing the vertical 

 scale into 10% bands. The time interval corresponding to the midpoint of 

 each band was. used as applicable to the 10% of the calls that fell within that 

 band. The midpoint values, the corresponding time intervals, and the 

 random numerals used are as follows: 



Based on the results indicated by Table III, that subscribers tend to make 

 repetitious attempts at a uniform pace or tempo, the time interval de- 

 termined by the numeral in the first place of a particular number of the 

 third set of random numbers was repeated each time that a particular call 

 was resubmitted. 



The results of the study of simulated traffic are as follows: 



Table V 



Trunks 

 Provided 



5 



6 



7 



8 



9 



10 



11 



12 



Attempts 



(Calls Offered 



Plus All 



Subsequent 



Attempts) 



1,658 

 1,287 

 1,147 

 1,071 

 1,027 

 1,011 

 1,005 

 1,000 



Overflows (Calls 



Encountering 



All Trunks 



Busies) 



720 



319 



155 



75 



28 



12 



5 







Ratios of 



Overflows to 



Attempts 



d " c + b 



.4343 

 .2479 

 .1351 

 .0700 

 .0273 

 .0119 

 .0050 

 .0000 



Approx. 



ch:s 



Handled 



ASOxe 



141 

 145 

 149 

 149 

 150 

 150 

 150 

 150 



