Frequency Domain Detection Method Based on WOLA Structure with Matched Filtering

At high sampling rates, the data throughput is large, and part of the data will be discarded because the data cannot be processed in real time, which leads to an increase in the probability of missing transitory signals. On the other hand, in practical test scenarios, the measured signal often has the characteristics of frequency domain sparseness, and the signal of interest is often located in a specific frequency band. Therefore, in this paper, a matching filtering frequency domain detection method based on the WOLA structure is proposed. The high-efficiency WOLA structure is used to decompose the signal according to different frequency bands. The output signal of each sub-frequency band is matched and filtered to extract the signal of interest from the noise. And the best judgment basis is provided at the detection time. According to the determined frequency domain position information, the system can filter out the signal of interest and prevent the risk of missing occasional signals, and solve the problem of high data throughput faced by transient signal processing in the time domain at high sampling rates. In the experiment, the pulse frequency hopping signal is used to verify the effectiveness, and the results show the effectiveness of the proposed detection method.