Hybrid Active Noise Reduction Headset based on fixed point processors

In this paper high load adaptive algorithms have been implemented on noise reduction headset to attenuate low frequency noise. By combining the characteristic of feedforward and feedback adaptive algorithms the attenuation of unwanted noise was done superlatively. By reducing program load due to fixed point calculations and omission of float calculations, high load algorithms are implemented on headset by one processor for both ear cups, in addition, by reducing program load the CPU operating frequency and its utilization percentage were decreased and therefore, the energy consumption was reduced. Experimental results show the high performance of the headset in low frequency noise and by reducing the CPU operation percentage the energy consumption was reduced over ۵۰ percent.

Keywords:

ANR Headset , Fixed-point DSP , Active noise control , Hybrid FXLMS