Abstract
The Modulated Wideband Converter (MWC) is one of the promising sub-Nyquist sampling architectures for sparse wideband signal sensing applications. Its frequency support detection and the reconstruction ability is well-defined through compressed sensing theory. However, the reconstruction performance of the MWC is strictly limited by the non-ideal components in practical implementations. Previous calibration methods exploit sequential single-tone signals to estimate the actual transfer function of the MWC, which is fairly time-consuming. We propose a new calibration method that estimates the actual sensing matrix coefficients of the MWC only with a single measurement based on a pilot multi-tone signal. The feasibility of the proposed calibration method is demonstrated in terms of normalized mean square error (NMSE) and image rejection ratio (IRR) of the reconstructed original signal.
Highlights
M ULTIBAND signal sensing has been successfully applied to cognitive radios and spectrum analyzers [1]–[3]
A sampling speed of current commercial analog-to-digital converters are often insufficient to capture such wideband signal based on Shannon-Nyquist theorem [4], [5] that states that the sampling speed must be at least twice higher than the bandwidth of the signal
We propose a novel calibration method for simultaneous estimation of all the components of the actual sensing matrix based on a single measurement with a nonsparse pilot signal
Summary
M ULTIBAND signal sensing has been successfully applied to cognitive radios and spectrum analyzers [1]–[3]. The previous works [7]–[10], [21]–[23] exploit the estimation method for actual sensing matrix based on a set of sequential measurements of single-tones with known frequencies. This iterative method is extremely time-consuming and the digital processing of the calibration turns into computationally complex due to the large number of calibration signals for real-time applications. We propose a novel calibration method for simultaneous estimation of all the components of the actual sensing matrix based on a single measurement with a nonsparse pilot signal.
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have