An Integration Optimized UWB Transmitted Reference System Based on Wavelet Denoising

This paper exploits the "localization" property of the wavelet transform to optimize the integration interval in transmitted reference (TR) system. The average correlation receiver is adopted in this paper to suppress the noise contained in the reference signal. The integration time optimization is based on the fact that the path energy decreases as the path delay increases and the tail of a frame time is dominated by noise, consequently, it is unnecessary for the correlation receiver to integrate over the full frame time. Hence, we apply the wavelet transform to locate the energy concentrated area and to optimize the integration time. It consists of two steps: the first one is the wavelet denoising which can intensify the signal part and reduce the noise component; the second one is to estimate the mean and covariance of multipath delay. Simulation results show that the approach performs better than the conventional TR and averaging only approach