A Total Least Squares Reconstruction Algorithm of UWB Signals Based on Sub-Nyquist Sampling
基于欠奈奎斯特采样的超宽带信号总体最小二乘重建算法

A sub-Nyquist sampling method is presented to reduce the ADC sampling rate in UWB wireless communications. Sampling rate of the proposed method is related to the signal innovation rate, which is about one tenth of the Nyquist rate in conventional Shannon sampling theorem. Fourier transform coefficients of the UWB signals are derived from theoretical analysis based on sub-Nyquist sampling. Then Total Least Squares (TLS) algorithm is proposed to estimate the parameters of the amplitudes and time shifts of impulse signals. The waveform of UWB signals can be reconstructed by convolving the estimated impulse signals with Gaussian monocycle. Simulation and experiment results show that the UWB signals can be accurately reconstructed, and the proposed methods outperform annihilating filter method in the presence of noise.