Rule-based noise cancellation: a fuzzy line enhancer

The extraction of a desired signal from extraneous noise is a typical research topic in the field of signal processing. The filter used to realize this task to enhance a narrow-band signal in the presence of broadband noise is called a line enhancer. In this paper, line enhancement (noise cancellation) by using fuzzy theory is introduced. By using a fuzzy logic approach, the locations of the lower bound and the upper bound of the main power band, in the spectrum of an arbitrary narrow-band signal that is corrupted by a broad-band noise, can be detected. The membership functions are described mathematically by sets of line equations based on analytic geometry rather than by conventional graphics and fuzzy inference is therefore mathematically implemented in software. A proper band-pass linear FIR filter is then automatically generated, based on the detected information to filter and finally enhance the signal of interest. The computer simulation, which is analyzed in this paper, demonstrates satisfactory results. The implementation considerations of the proposed fuzzy line enhancer are compared with those of the existing adaptive line enhancers; it is satisfactory, and the algorithm is simpler.