A Wideband Power Spectrum Estimation Method Based on Multi-Channel Compressed Sampling

Abstract

The invention relates to a broadband power spectrum estimation method based on multi-channel compressed sampling. The method is oriented to broadband signals, and the analog front-end adopts a multi-channel compressed sampling method to perform discrete Fourier transform on the sampling value of each channel. First, the time domain The multiplication is converted into extraction after convolution, and then converted into frequency domain multiplication, and the cross-power spectrum relationship between each channel is established. The power spectrum of the signal is estimated using the direct method, where the equations are solved using the method of least squares. The present invention has the advantages of low sampling rate, low computational complexity, etc., and does not need a priori information of signal frequency domain sparsity.

Description

technical field [0001] The invention relates to the technical field of communication signal processing, in particular to a wideband power spectrum estimation method based on multi-channel compressed sampling. Background technique [0002] With the application of various wireless communication services, the demand for high-speed data communication makes spectrum resources become scarce, but the existing wireless communication systems all adopt a fixed channel allocation strategy, which makes the spectrum utilization rate low. The cognitive radio technology using the dynamic access method makes the spectrum fully utilized. One of its key technologies is spectrum sensing, which identifies unused frequency bands in a large frequency range for dynamic wireless use by unlicensed secondary users. Access to improve the utilization of the actual spectrum. [0003] According to the Nyquist sampling law, in order to be able to perceive the entire broadband signal, the sampling frequen...

AI Technical Summary

Problems solved by technology

[0005] However, the current spectrum sensing algorithm based on compressed sampling has the following problems: 1). Using convex optimization, greedy algorithm and other methods to achieve signal recovery is often extremely complex, resulting in long spectrum sensing time and high hardware overhead

However, in a dynamically changing spectrum access environment, it is difficult to accurately obtain the sparsity of the signal to be sensed in advance

3). If there are many users connected, the perceived signal is non-sparse or dense spectrum signal. At this time, the performance of the method of restoring the original signal based on compressed sampling will deteriorate sharply, or it will no longer be applicable.

Images