An improved variable step size LMS algorithm
Xueli Wu, Liang Gao, Zizhong Tan
Abstract
Xueli Wu, Liang Gao, Zizhong Tan
Abstract
The traditional variable step size least mean square (LMS) algorithm has many weakness, such as poor convergence speed and susceptible to noise interference. In order to improve the performance of the algorithm, an improved variable step size LMS algorithm is presented, and it is applied to the noise cancelling system of communication. By using the hyperbolic tangent function, the relationship between step size and error signal is established. In this algorithm, the step size factor is adjusted by the absolute value of the product of the current and former errors. The algorithm also introduces the disturbance of the absolute estimation error to update the tapping vector of the self-adaptive filter. The simulation results show that the proposed algorithm has a faster convergence speed than traditional LMS algorithm and CTanh-LMS algorithm. And it obtains a good results in the noise cancelling system of communication.
OpenAlex reports 8 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
The traditional variable step size least mean square (LMS) algorithm has many weakness, such as poor convergence speed and susceptible to noise interference. In order to improve the performance of the algorithm, an improved variable step size LMS algorithm is presented, and it is applied to the noise cancelling system of communication. By using the hyperbolic tangent function, the relationship between step size and error signal is established. In this algorithm, the step size factor is adjusted by the absolute value of the product of the current and former errors. The algorithm also introduces the disturbance of the absolute estimation error to update the tapping vector of the self-adaptive filter. The simulation results show that the proposed algorithm has a faster convergence speed than traditional LMS algorithm and CTanh-LMS algorithm. And it obtains a good results in the noise cancelling system of communication.
Key concepts: Algorithm, Least mean squares filter, Computer science, Adaptive filter, Noise (video), Variable (mathematics), Convergence (economics), Rate of convergence