Multi-FFT Signal Processing for Unknown-Band Low-SNR Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems struggle to accurately detect signals with unknown frequency bands due to increased noise and deteriorated S/N ratio when setting the observation band to broadband, making it difficult to identify unknown signals within received broadband signals.
Innovation Solution
A signal processing device that includes an A/D converter, FFT processing, correlation calculation, and data sequence selection to estimate data sequences with unknown frequency bands by converting received signals into time-domain sequences, performing FFT processing with multiple FFT points, calculating correlations between frequency-domain sequences, and selecting sequences with correlations above a threshold.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the observation band is set to broadband to detect signals with unknown frequency bands, then the coverage of detectable signals is improved, but the noise increases and the S/N ratio deteriorates
Solution Approach 1:
The patent segments the broadband signal into multiple narrowband signals by performing FFT processing with different FFT points. Each narrowband signal corresponds to a specific frequency range, allowing the system to maintain the versatility of broadband monitoring while achieving the signal quality of narrowband processing through selective correlation analysis.
Solution Approach 2:
The patent changes the parameter of FFT points to create multiple narrowband signals from the broadband input. By varying the FFT point parameter, the system generates frequency-domain data sequences with different resolution characteristics, enabling adaptive filtering and correlation calculation to improve S/N ratio while maintaining broadband coverage capability.
2Measurement precision
If broadband signals are converted into frequency domain data sequence, then the frequency analysis capability is improved, but it becomes difficult to identify unknown signals in the data sequence due to noise
Solution Approach 1:
The patent introduces an intermediary correlation calculation process that operates on the frequency-domain data sequences. By calculating correlation between sequences obtained through FFT with different points, the system creates an intermediate representation that highlights consistent frequency components across different transforms, effectively filtering noise while preserving unknown signal identification capability.
3Productivity
If all received broadband signals are transmitted to ground station for processing, then the processing capability is improved, but the transmission capacity requirement increases
Solution Approach 1:
The patent extracts only the necessary frequency-domain data sequences from the broadband signal that show significant correlation above a threshold value. This extraction process identifies and isolates potential unknown signals while discarding noise components, allowing transmission of only essential data to the ground station rather than all received broadband signals.
Data Source
AI summary
To provide a signal processing device or the like that can estimate a data sequence including a signal with an unknown frequency band from among received signals having a small S/N ratio.A signal processing device 2A includes an A/D converter 215 that converts signals received by the flying object over a predetermined period of time by varying a time into a time-domain data sequence for each time, a fast Fourier transform (FFT) processing device 221 that performs FFT processing on the time-domain data sequence for each time with a plurality of different FFT points and outputs a frequency-domain FFT result data sequence at each FFT point, a correlation calculation device 222 that calculates a correlation between the frequency-domain FFT result data sequences at different times with the same FFT point for each segment based on a frequency resolution unit based on a sampling frequency of the A/D converter and the same FFT point and that performs the calculation for each of the plurality of different FFT points, and a data sequence selecting device 223 that selects a data sequence having the correlation equal to or greater than a threshold value.