Comparison of detection performance of a time domain and frequency domain adaptive algorithm

The detection of narrowband tonals amid broadband noise is typically performed by an FFT processor followed by square law detection and integration. The use of an adaptive Time Domain filter to augment the processor has been shown to provide potential robustness in the processor for the case of non-ideal broadband background noise. In this paper, the performance of an alternate frequency domain augmented narrowband processor is compared to the Time Domain adaptive filter. The detection performance of the two different augmented processors is measured by means of Receiver Operator Characteristic (ROC) curves. It is shown that the detection performance for the two augmented processors is approximately the same for the case of large post detection time-bandwidth products.