The normalized fractionally-lower order moment (FLOM) algorithm exhibits fast convergence but high excess mean squared error (MSE) when the order is less than 2. This paper proposes a method using variable order moments to adaptively changing the order during adaptation, thus achieving both fast initial convergence and low excess MSE in the steady state. The algorithm is applied to both Gaussian and heavy-tailed non-Gaussian clutter suppression in phased array applications. The results show better performances of the proposed algorithm over the normalized FLOM and normalized Least Mean Square (NLMS) algorithms. The proposed algorithm also performs well in other adaptive filtering applications such as system identification and noise/echo suppression.
Y. R. Zheng, "A Variable-Order FLOM Algorithm for Heavy-Tailed Clutter Suppression," Proceedings of the IEEE Military Communications Conference, 2008. MILCOM 2008, Institute of Electrical and Electronics Engineers (IEEE), Nov 2008.
The definitive version is available at https://doi.org/10.1109/MILCOM.2008.4753305
IEEE Military Communications Conference, 2008. MILCOM 2008
Electrical and Computer Engineering
United States. Air Force. Office of Scientific Research
Keywords and Phrases
Echo Suppression; Mean Square Error Methods; Phased Array Radar; Space-Time Adaptive Processing; Adaptive filters
Article - Conference proceedings
© 2008 Institute of Electrical and Electronics Engineers (IEEE), All rights reserved.
01 Nov 2008