Fast approximated power iteration subspace tracking
- 18 July 2005
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Signal Processing
- Vol. 53 (8) , 2931-2941
- https://doi.org/10.1109/tsp.2005.850378
Abstract
This paper introduces a fast implementation of the power iteration method for subspace tracking, based on an approximation that is less restrictive than the well-known projection approximation. This algorithm, referred to as the fast approximated power iteration (API) method, guarantees the orthonormality of the subspace weighting matrix at each iteration. Moreover, it outperforms many subspace trackers related to the power iteration method, such as PAST, NIC, NP3, and OPAST, while having the same computational complexity. The API method is designed for both exponential windows and sliding windows. Our numerical simulations show that sliding windows offer a faster tracking response to abrupt signal variations.Keywords
This publication has 32 references indexed in Scilit:
- Adaptive ESPRIT algorithm based on the PAST subspace trackerPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2004
- Sliding Window Adaptive SVD AlgorithmsIEEE Transactions on Signal Processing, 2004
- Numerically-robust adaptive subspace tracking using Householder transformationsPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2002
- Fast adaptive eigenvalue decomposition: a maximum likelihood approachSignal Processing, 2002
- Efficient, high performance, subspace tracking for time-domain dataIEEE Transactions on Signal Processing, 2000
- Fast recursive subspace adaptive ESPRIT algorithmsIEEE Transactions on Signal Processing, 1998
- Fast algorithms for updating signal subspacesIEEE Transactions on Circuits and Systems II: Analog and Digital Signal Processing, 1994
- An updating algorithm for subspace trackingIEEE Transactions on Signal Processing, 1992
- Noniterative subspace trackingIEEE Transactions on Signal Processing, 1992
- Matrix pencil method for estimating parameters of exponentially damped/undamped sinusoids in noiseIEEE Transactions on Acoustics, Speech, and Signal Processing, 1990