A Recursive Least P-Order Algorithm Based on M-Estimation in a Non-Gaussian Environment

• Published in: IEEE transactions on circuits and systems. II, Express briefs Vol. 70; no. 8; pp. 2979 - 2983
• Main Authors: Wang, Gang, Ouyang, Linqiang, Tong, Linling, Fang, Qiang, Peng, Bei
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.08.2023; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Adaptive algorithms; Adaptive filtering; Adaptive filters; Cost function; Estimation; Filtering; Filtering algorithms; Kernel; Logistics; M-estimation; Random noise; recursive least p-order; Robustness (mathematics); Steady-state
• ISSN: 1549-7747; 1558-3791
• DOI: 10.1109/TCSII.2023.3253500

In recent years, adaptive filtering algorithms based on M-estimation can effectively handle non-Gaussian noise, but large outliers still exist which will seriously disrupt the performance of the algorithm. Here, the recursive least p-Order algorithm based on M-estimation (MRLP) is proposed. The MRLP combines the advantages of M-estimation and the <inline-formula> <tex-math notation="LaTeX">l_{p} </tex-math></inline-formula> norms. The proposed MRLP can mitigate the impact of large outliers and exhibit strong robustness when suffering from non-Gaussian noise, and can also be well applied to tracking signals. Numerical simulations verify the excellent performance of the proposed algorithm.