Where binary code imbalance decreases DC bias sensitivity

Abstract

The imbalance of a binary code with elements either +1 or -1 is the number of -1 elements subtracted from the number of +1 elements. A balanced code is one with 0 imbalance. Balanced codes are often used for pulse compression because the DC response of a linear pulse compression module is proportional to the imbalance. A nonzero DC bias combined with an unbalanced code produces a nonzero output. In a radar signal processor, this can mean susceptibility to false alarms from signals that alias to DC at pulse compressor input. This paper examines a common radar signal processor subchain for which some imbalance in pulse compression codes can reduce bias sensitivity. The chain consists of R/theta limiter, pulse compressor and I/sup 2/+Q/sup 2/ modules. It is shown that for an input consisting of DC bias and additive Gaussian noise, there is a nonzero value for the code imbalance for which the mean and variance of the output are insensitive to the magnitude and phase of the input bias.