Static and Dynamic Signal Processing Methods for Noise Cancellation in Sound and Electromagnetic Environment

摘要

The observed phenomena in actual sound and electromagnetic environment are inevitably contaminated by the background noise of arbitrary distribution type. Therefore, in order to evaluate sound and electromagnetic environment, it is necessary to establish some signal processing methods to remove the undesirable effects of the background noise. In this paper, we propose noise cancellation methods for estimating a specific signal with the existence of background noise of non-Gaussian distribution from two viewpoins of static and dynamic signal processing. By applying the well-known least mean squared method for the moment statistics with several orders, practical methods for estimating the specific signal are derived. The effectiveness of the proposed theoretical methods is experimentally confirmed by applying them to estimation problems in actual sound and magnetic field environment.