Digital Matched Filtering of Arbitrary Spread-Spectrum Waveforms using Correlators with Binary Quantization

Optimum processing of spread-spectrum waveforms in the presence of white Gaussian noise is possible by using matched filters. The present paper deals with digital matched filters (DMF) which apply digital 1-bit correlators and utilize a binary quantization scheme called 4-phase quantization. The application of 1-bit correlators reduces technical expense in comparison with digital filter designs using a finer amplitude quantization. On the other hand binary correlation entails a noticeable degradation of S/N performance in comparison with analog or quasi-analog devices. In our paper the concept of the so-called 4-Phase DMF and an expression for its S/N degradation are briefly presented. This expression will be discussed and general criteria for optimum filter implementation and waveform design with respect to as low a degradation as possible will be derived. It will be shown that for arbitrary constant envelope spread-spectrum waveforms the S/N degradation can be kept below 2.87 dB by proper selection of the binarily quantized reference waveform programmed into the DMF. For some specific spread-spectrum waveforms such as PN-MSK, PN-PSK and PN-QP SK the lower bound 1.96 dB on the degradation of the 4-Phase DMF can be reached. The paper is supplemented by the description of a DMF which is presently under construction.