Online secondary path estimation plays an important role in design of
active noise control (ANC) systems and the final achievable performance for the overall system. However, due to cross coupling and the effect of nonlinearities in the system, the results are usually poor, and the problem requires specific attentions. In this study, we aim to investigate the impact of finite measured multi-sine data as an input excitation signal to find the best linear approximation for the secondary path estimation.
The approach relies on robust local polynomial method as an advanced technique for non-parametric measurement of the room impulse responses. In this way, it is possible to quantify various sources of uncertainties such as measurement noise, leakage, and nonlinear distortions by deeper analysis of the results. The information can be used to understand the situation of ANC system and improve the online the secondary path estimation, subsequently.