Digital signal processing for optimum wideband channel estimation in the presence of noise

摘要

The problem of estimating the impulse response of a transmission channel in communication systems has great practical relevance. The performance of digital signal processing algorithms suitable for channel estimation with wideband sounding signals in the presence of white and nonwhite noise are evaluated and compared. The algorithms are based on conventional matched filtering, which optimises the signal-to-noise ratio at the receiver output, and optimum unbiased estimation, which totally avoids the harmful systematic errors comparable to the correlation sidelobes at the output of a matched filter. The results presented can be used in an actual application to find the most suitable combination of channel estimation algorithm and spreading code with respect to a specific noise configuration. Simulation results based on different noise configurations give an indication of the performance of the different digital channel estimation algorithms in the presence of white and nonwhite noise.