Influence of signal level quantization at the input of a beam-forming system on the noise immunity of the receiving section

The influence of signal quantization on the noise immunity of a beam-forming system for spatial signal processing in a narrowband Gaussian field of reverberation noise, isotropically distributed random noise, and local noise is analyzed. Quantitative estimates of the noise immunity for the detection of a simple signal are obtained for a random noise field, a field comprising isotropic and local noise components, and a field comprising isotropic random and reverberation noise components. The directivity patterns of the system are determined with digital representation of the input information for various numbers of bits, various ratios between the local noise and signal levels, and various angles of arrival of the local noise. It is concluded that the number of bits used to represent the input information must be of the order of 6-8 and that, in order to ensure the optimal quantization step, the processing system must have the capability of measuring the rms value of the total noise and adjusting the noise level accordingly at the analog to digital converter input in correspondence with the nature of the noise.