Multichannel Load Calculation Using the Monte Carlo Method

This paper describes a technique for calculating the multichannel loading of telephone carrier systems. The approach described is designed to exploit the modern computer more effectively while, at the same time, avoiding some of the weaknesses encountered in previous calculations, namely: (1) the dependence on a "multichannel peak factor" (MCPF), which is established experimentally by partially justified methods (2) the addition of a MCPF for a fixed number of channels to the multichannel equivalent-level distribution (1 percent point) to obtain the instantaneous multichannel peak value. Ideally, this MCPF should be allowed to vary simultaneously with the channel activity and other variables. A further advantage, in the approach submitted here, is that the results obtained contain additional information, derived from the fact that the multichannel load is given in terms of a distribution of peaks with specified probability values. This information is required when dealing with overload and near-linear distortion noise analysis, or in studies involving PCM companding laws. Briefly, the method used is to derive a probability distribution function which describes the amplitude characteristic of multichannel signals as functions of the number of channels in the system. Two basic sets of data are used: (1) the channel activity (2) the instantaneous amplitude probability distribution for an active channel. This information is used in conjunction with the Monte Carlo sampling technique to simulate the instantaneous amplitude from the multichannel signal. This simulation process, incorporated into a computer program, provides a means of generating samples and, by using a sufficient number of samples, a probability distribution is obtained to describe the multichannel loading.