STABILITY ANALYSIS OF STEPWISE SIMULTANEOUS PERTURBATION STOCHASTIC APPROXIMATION FOR ACTIVE NOISE CONTROL

Active noise control (ANC) is particularly effective in mitigating low-frequency noise, with its primary objective being to achieve maximum noise reduction while minimizing both measurement and computational costs. The filtered-x least mean square (FxLMS) algorithm has been widely adopted due to its effective balance between performance, ease of implementation, and computational efficiency. However, FxLMS relies on accurate secondary path estimation, which can be computationally intensive and time consuming. Recently, as a model-free method, the simultaneous perturbation stochastic approximation (SPSA) algorithm has been successfully applied in broadband, multi-channel, and time-varying environments. Despite its demonstrated effectiveness in noise reduction, the stability of SPSA has received limited attention. Following a recently proposed stepwise SPSA extension, which significantly enhances the stability of classical SPSA, this work investigates its stability characteristics with respect to key algorithmic parameters. The analysis provides insights into the parameter selection process, enabling efficient noise reduction while ensuring stability.