330 THE BELL SYSTEM TECHNICAL JOURNAL, MARCH 1953 



The chief difference is that prehminary decisions concerning the size 

 and general characteristics of the office to be tested will be made. For 

 example it may be decided to test an office in the twenty line link frame 

 size range serving mixed business and residential subscribers with a 

 high percentage of calls completed within the office. Based on general 

 knowledge of subscriber behavior, the number of subscribers necessary 

 to present a suitable traffic load to this number of link frames will be 

 determined. Values for average holding time and calling rate associated 

 with these subscribers must also be developed. These may, in part, be 

 determined from estimates or specific knowledge of the traffic capacity 

 of the line link frames, taking care that the figures are typical of such a 

 group of subscribers as determined by field observations. Since we are 

 usually concerned with determining the maximum safe capacity of the 

 system, full load or overload conditions will be assumed and the usual 

 margins for future growth considered in engineering an actual office 

 will be omitted. Decisions will also be made as to the number of other 

 offices to which this office has trunks and the percentages of the total 

 traffic originated and terminated in each of these offices. 



Having made these preliminary assumptions concerning the nature 

 and environment of the office to be tested, the office is then engineered 

 according to the best available information. The numbers of registers 

 and markers are determined and arranged in connector groups accord- 

 ing to standard procedures. The sizes of the trunk groups to various 

 connecting offices are determined and the placement of trunks on the 

 trunk link frames chosen according to the usual practice. All similar 

 factors concerning quantities and arrangement of equipment are deter- 

 mined. The throwdown machine is then set up to simulate this office. 

 This will involve crossconnections in the gate circuits and arrangement 

 and designation of the facilities provided for keeping the busy-idle records 

 of such items as lines, trunks, registers and links. 



The second phase in the preparation is to produce data representing 

 calls presented to the system during the time interval to be studied. 

 This is accomplished by choosing a random number for each call. This 

 number must contain a sufficient number of digits to specify all the 

 pertinent data necessary to describe the call. These digits are assigned 

 to represent certain factors. For example, the first six digits represent 

 the time of origination. The next two digits specify the type of call. 

 This is done on a percentage basis. For example, if 40 per cent of the 

 traffic is to be locally completed, the numbers 00 through 39 in these 

 places would indicate an intraoffice call, if 25 per cent is to be outgohig to 

 other offices the numbers 40 through 64 would indicate this type of 



