Research On Different Filtering Algorithms Based On Adaptive Noise Cancellation System

Adaptive noise cancellation system is offset by calculation of the noise, so as toobtain the useful signal’s system. The adaptive filter is the main component of theadaptive noise cancellation system. And adaptive filtering algorithm is the core workof the adaptive filter. Therefore, the scope of research work in this paper and theobject are different adaptive filtering algorithm based on the adaptive noisecancellation system. The method of the paper is different adaptive algorithm anddifferent LMS algorithm-combined to achieve the adaptive filter to the noise signal’sfiltering. The conclusion is that the paper proposes variable step-size LMS algorithmand different LMS algorithm-combined to achieve the adaptive filter of the noisesignal, which the filtering effect is better than the other filtering algorithm.Firstly, the paper analyzed adaptive filter’s basic principle, structure and theapplication. Secondly, the paper introduced the adaptive filtering algorithm. Adaptivefiltering algorithm included the fixed step size LMS algorithm, the normalized LMSalgorithm and variable step size LMS algorithm. According to the variable step-sizeidea, the paper proposed a variable step-size LMS algorithm based on Bessel function.Adaptive noise cancellation technology’s simulation of variable step size LMSalgorithm based on the cosine function and the paper’s variable step-size LMSalgorithm were simulated in the MATLAB environment. The simulation resultsshowed that the paper’s algorithm can achieve faster convergence speed, and thesmaller disorders, and better tracking performance, which could remove theinterference of uncorrelated noises effectively. It can also maintain a smallersteady-state error when it was inputted a mutation signal. At present, people haveproposed various variable step-size LMS algorithms. But these variable step-sizeLMS algorithms were only applied to the one-level adaptive filters. Therefore, thepaper also proposed the multi-level filter, which the design idea of Multi-level filterwas made of different algorithms of the adaptive filter to be combined together toform a filter group. There were the combination of two fixed step-size LMS algorithm, the combination of fixed step-size LMS algorithm and variable step-size LMSalgorithm, the combination of variable step-size LMS algorithm and variablestep-size LMS algorithm. Lastly, the paper presented a new adaptive algorithm andthe multi-level filter to achieve the adaptive filter to the noise signal’s filtering in theMATLAB environment. Theory analyzing and simulation effects showed that thenormalized LMS algorithm filtering result is better than the fixed step size LMSalgorithm, and variable step size LMS algorithm filtering is better than thenormalized LMS algorithm filtering, and the different LMS algorithm-combined forthe two level filtering is obviously better than the single fixed step size LMSalgorithm and NLMS algorithm for the filtering. The two level filtering improved thefilter performance in terms of the fast convergence speed, tracking speed and the lowmaladjustment error.