Adaptive filtering for noise cancellation and signal analysis in real-time

This paper presents advanced digital signal algorithms for adaptive filtering applied for noise cancellation and signal analysis in real-time. Correlated and not-correlated signal parts are distinguished by such methods based on the statistical characteristics of the received signals. Real-time measurements of electromagnetic interference (EMI) in noisy environments are shown. An advanced digital signal processing technique for fast measurements of electromagnetic interferences is presented, carrying out the measurements in time-domain and reducing the total measurement time by orders of magnitude. For optimum noise suppression performance, different approaches of adaptive filter algorithms were investigated and enhanced with respect to implementation of the noise cancelling algorithm on field programmable gate arrays (FPGA) and enabling signal processing in real-time. Noise suppressing in real-time is shown in frequency bands of over 125MHz at once. Several real-time bands up to 1 GHz are measured and signals completely covered by ambient noise could be successfully detected, e.g. in the DAB, the DVB-T and LTE frequency bands.