Multi-channel adaptive feedforward systems for multi-input multi-output active control of broadband road noise

Multi-input, multi-output (MIMO) active noise control (ANC) has a wide variety of applications for low-frequency noise control configured with the standard filtered-x least mean squares (FxLMS) algorithm. However, the conventional active control of broadband noise, such as vehicle road noise control using multiple reference accelerometers, multiple control speakers and error sensors, is often limited in practice by slow convergence speeds, high computational demand and poor overall noise reductions. These problems might be overcome by the use of subband adaptive filtering. In a simple context of multi-channel ANC system, this paper presents two MIMO ANC systems based on the FxLMS algorithm and the delayless subband FxLMS (DSFxLMS) algorithm, respectively. Two implementations of the secondary path modeling are introduced. Numerical simulations using synthesized road noise data and secondary path models demonstrate that the MIMO DSFxLMS algorithm achieves better overall noise cancellation as compared to the MIMO FxLMS algorithm.