Implementation of magnitude estimation algorithm for hearing aid

The hardware implementation of a magnitude estimation algorithm for a hearing aid is presented in this paper. Digital signal processing (DSP) is used in almost all the modern digital hearing aids available today. The processing of the sound signals is carried out in the frequency domain. Fast Fourier transform (FFT) converts the time domain signal to frequency domain and gives a complex number output representing each frequency component in the audible frequency spectrum. The magnitude estimation block estimates the magnitude of these complex numbers. This estimated magnitude is used as an input to apply various DSP algorithms for sound processing in the hearing aid. In this paper the Alpha max-Beta min and the coordinate rotation digital computer (CORDIC) algorithms for magnitude estimation of a complex number are compared from the hardware implementation point of view. The emphasis of this paper is the hardware implementation of the CORDIC algorithm in application specific integrated circuit (ASIC) design flow and its performance analysis. The CORDIC algorithm is implemented because of its accuracy and implementation simplicity. The paper also presents the simulation and synthesis results obtained for hardware implementation of four iterations of the CORDIC algorithm for magnitude estimation.