Further Analysis of Errors Reported in 'Capabilities of the Telephone Network for Data Transmission'

The purpose of this paper is to describe certain characteristics of errors affecting digital data in the switched telephone network. The present results are based on an analysis of error data recorded in a field testing program conducted by the D a t a Transmission Evaluation Task Force of the Bell System. These data are summarized in a previous paper by Alexander, Gryb, and Nast 1 which sets forth numerous conclusions about the telephone plant, based on the test program. The present paper reports the results of a sequence of statistical investiga1399 1400 T H E BELL SYSTEM TECHNICAL J O U R N A L , JULY 1962 tions which have made it possible to classify most of the errors into several populations. The effect of the test equipment on the time-distribution of errors is also discussed. I I . T H E TASK FORCE TESTING PROGRAM This section contains a description of those aspects of the test program which bear on the present discussion. Test calls were made by the Task Force over a wide variety of circuits. Both local and long distance calls were made over the switched telephone network and the performance of the circuits was recorded. Data concerning 1010 test calls were gathered and used for analysis. The test calls were of three categories: 10-minute calls at 600 bits/second, 10-minute calls at 1200 bits/second, and 30-minute calls at 1200 bits/second. Each of these categories was further subdivided into "Exchange Calls" which used no long distance switching facilities, "Short Haul Long Distance Calls" which were made over distances of less than 400 miles, and "Long Haul Long Distance Calls" which were made over distances of between 400 and 3000 miles.