Bark Frequency Spectrum in Parallel-form Remote Active Noise Control

Active noise control (ANC) is widely used to reduce low-frequency noises. ANC aims to cancel the noise by generating another noise with equal magnitude and opposite phase. The superposition of the noise and anti-noise cancel out each other to create a zone of quiet (ZoQ) at an error microphone. However, in case of physical limitations or application constraints to place the error microphone, the virtual sensing algorithms are used to cancel the unwanted noises at the desired location. In many practical ANC applications, the primary noise contains multiple discrete low frequencies. A narrowband active noise control structure is often applied to reduce unwanted noise when the multiple tones have proximity frequencies. In this paper, a parallel-form remote ANC algorithm is proposed to cancel the narrowband noises at the remote location. A parallel-form ANC structure separates the discrete frequencies into a series of adaptive filters. Then delayless bandpass filter bank is used to split the measured error signal, and use the individual error signal to update the corresponding adaptive filters. The transfer function of the remote controller is derived in the acoustic domain, which moves the ZoQ from the error microphone to the nearby remote location. Computer simulations are performed to verify the performance of the proposed algorithm.