Analysis of Adaptive Filter Algorithms in Real Time Signals

Noise reduction in Real time signal is a key challenge in Communication system. Active noise cancelling technique is used to minimize the distracting background noise from the real-time signal and to provide pleasant voice during communication in portable devices. Active Noise Cancellation is a technique of reducing unwanted sound by generating the anti-noise signal. The anti-noise signal is generated by using algorithms like Least Mean Square Algorithm (LMS), Normalized Least Mean Square Algorithm (NLMS), and Filtered X Least Mean Square Algorithm (FXLMS) for developing the ANC system. The generated anti-noise helps to reduce the distracting background noise and provides noise free desired signal output. In the proposed work anti-noise is generated for various real-time recorded noise signals like crow noise, train whistle noise, AC noise and multiple noises obtained as the combination of two noises using LMS, NLMS and FXLMS algorithm. The anti-noise signal generated is then used to cancel the respective noise added with the different input signals. The input signals considered are sine signal, real-time voice and song. Performance analysis of the algorithms is carried out based on the parameters like RMSE, SNR, and convergence time using MATLAB simulation software.