On evaluating polarity-coincidence correlation when the two inputs are statistically dependent (Corresp.)

Polarity-coincidence correlation (pcc) is usually analyzed under the assumption of independent noise inputs and small input-signal power. Thus the difficulty of evaluating the variance of the PCC statistic for inputs with arbitrary cross correlations is avoided. An expression for the variance that can be conveniently evaluated on a computer is discussed. As one example, the PCC statistic is analyzed for a strong Markovian signal that is added to two independent, white noise inputs. The effect of hard limiting is determined as a function of the input signal-to-noise ratio. In another example, the PCC detector is analyzed for a small Markovian signal that is added to two dependent Markov noise inputs. In this case, the cost of clipping increases substantially with the noise correlation.