Multi-channel adjoint least mean square algorithm with momentum factor applied on active noise control

Active noise control (ANC) is extensively utilized to attenuate unwanted environmental noise, creating a more conducive environment for work and daily activities. Traditional approaches face challenges when scaled to larger areas using multi-channel ANC (McANC) due to escalating computational burden and slower convergence speeds as channel numbers increase. To overcome these limitations, we introduce a multi-channel adjoint least mean square algorithm with a momentum factor (Mom-McALMS) designed to achieve optimal control at steady-state while reducing computational demands and accelerating convergence. Furthermore, the theoretical analysis presented in this paper reveals the impact of the step size and momentum factor on the stability of the proposed method. Numerical simulations validate the theoretical findings and the improved noise reduction performance of the proposed Mom-McALMS for both stationary and time-variant noises. Furthermore, real-time experiments conducted on a McANC window validate the effectiveness of Mom-McALMS in attenuating noise and improving convergence speed when dealing with different types of real-world noises.