An efficient extension of the zero-crossing technique to measure frequency of noisy signals

The paper contains a proposal for the estimation of the frequency of a sine signal corrupted by zero-mean additive white gaussian noise. The method resorts to the histogram of the time intervals between adjacent zero-crossing samples. Validation has been carried out for various noise levels and results show that the relative error is much less than other algorithms for frequency estimation. Performances are also compared to those of numeric counters. The method can be easily implemented of low cost devices and is computationally very efficient.