Spectrum analyzer overlap requirements and detectability using discrete Fourier transform and composite digital filters

In spectrum analysis, a primary measure of performance is the minimum detectable signal (MDS). Detectability dictates the compute speed necessary to meet both time and frequency overlap requirements. It is becoming increasingly popular to compute the discrete Fourier transform (DFT), and these elemental filters are complex weighted or otherwise combined to form more complex filters. Applications of this include the proportional bandwidth filter comb and all time‐weighting functions such as Hanning, Hamming, and Taylor or their frequency equivalents. Overlap requirements are thereby changed and MDS must be defined so it can be extrapolated to these composite filters. The overlap requirement and MDS are so defined in this paper. Experimental performance is then given and compared to the predictions.