The Transmission Performance of Bell System Intertoll Trunks

The noise waveforms found on voice bandwidth telephone channels are generally recognized to be non-gaussian in their amplitude distribution. This paper presents data which suggests that a simple exponential is a good function to describe amplitude densities in the extreme tails.A comprehenisve model of impulse noise as viewed on trunk groups is then presented. The model relates the distributions of impulse noise levels and impulse noise ωυηΐ8. r. INTRODUCTIONNoise on telephone channels has been measured for years with in struments which are constructed to enable reasonably good correlations between the reading obtained and the annoyance of the noise during a telephone conversation.' Fluctuations of the meter pointer during a measurement are either ignored or mentally averaged by the observer, depending upon their frequency of occurrence and their magnitude. With the introduction of data transmission on the telephone network, the relatively frequent high amplitude excursions of the noise waveform were viewed as a "new" kind of noise, primarily because they were generally not annoying in voice communication and it was recognized that no meaningful measure of them could be obtained with the stand ard noise measuring sets. The term "impulse noise" was applied to these high excursions and new instruments were designed to measure them.* The significance of impulse noise in data transmission has given rise to a great deal of effort devoted to its measurement, characterization, and evaluation as a transmission impairment.'"* (For an extensive bibliography, see Ref. 3.) Several models have been suggested to de scribe the erratic behavior and clustering phenomena associated with 3243

Amplitude Distributions of Telephone Channel Noise and a Model for Impulse Noise

Fennick¹

1969

1969-70 Connection Survey: Analog Transmission Performance on the Switched Telecommunications Network

Duffy¹,

Thatcher²

1971

To characterize the transmission performance of the Bell System switched telecommunications network, Bell Telephone Laboratories conducted a survey of toll connections during 1969 and 1970. Connections were established between Bell System end offices chosen by statistical sampling techniques. Both analog and data transmission tests were performed. A summary of analog transmission performance is presented in this paper. It contains estimates for noise, loss, attenuation distortion, envelope delay distortion, peak‐to‐average ratio, frequency offset, level tracking, nonlinear distortion, and phase jitter for toll calls within the Bell System. Accompanying papers discuss data transmission error performance at various speeds between 150 and 4800 bits per second.

The Transmission Performance of Bell System Toll Connecting Trunks

Kessler¹

1971

A systemwide survey of the performance of Bell System toll connecting trunks was undertaken in 1966. Various results of this survey have been used in other studies since that time. This paper collects the main results in a single writing. The sampling plan and measurement procedures are discussed briefly. Measurement results are presented in distributional form and as estimates for the transmission parameters relating to loss, message circuit noise, impulse noise, relative envelope delay, P/AR, harmonic distortion, and physical characteristics. Some results are presented separately for major facilities and facility categories.