An Improved Adaptive Subspace Tracking Algorithm Based on Approximated Power Iteration

Wu, Qiuji; Zheng, Jiansheng; Dong, Zhicheng; Panayırcı, Erdal; Wu, Zhiqiang; Qingnuobu, Ren

doi:10.1109/access.2018.2863557

Cited by 8 publications

(6 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…δh k = 1 K E δh h h k [n] 2 2 and = (C C C H ξ g C C C ξ g ) −1 C C C H ξ g C C C ξ ξ C C C ξ g (C C C H ξ g C C C ξ g ) −H . Here, we have also used the following: 11) and the fact that…”

Section: Discussionmentioning

confidence: 99%

“…Therefore, [33], [34] have proposed to estimate the column space of α using the PAST subspace tracking algorithm and its extensions. To this end, one usually considers the following model for z z z The use of the LS criterion in (11) implicitly assumes that the additive noise η η η[n] is spatially white. However, in the context of system identification, the additive noise v v v α [n] may not be spatially white.…”

Section: B Recursive Estimation Of a A A And C C Cmentioning

confidence: 99%

“…Most RSI algorithms estimate the unknown state matrix A A A and the output matrix C C C from the ''data equation'' consisting of the input and output samples of the systems. This requires the estimation of the subspace of the extended observability matrix, which requires SVD [10] or the use of more efficient subspace tracking algorithms [11], [12]. The computation of the state matrix A A A also requires the evaluation of a pseudoinverse and few works discussed the online estimation of the model order since it involves the computation of its eigenvalues.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

A Variable Regularized Recursive Subspace Model Identification Algorithm With Extended Instrumental Variable and Variable Forgetting Factor

2020

View full text Add to dashboard Cite

This paper proposes a new smoothly clipped absolute deviation (SCAD) regularized recursive subspace model identification algorithm with square root(SR) extended instrumental variable (EIV) and locally optimal variable forgetting factor (LOFF). The proposed algorithm is based on a new local polynomial modeling (LPM)-based variable FF EIV projection approximation subspace tracking (PAST) and multivariate recursive least squares algorithms for estimating respectively the subspace of the extended observability matrix, and the input matrix and feedthrough matrix. The SREIV-PAST algorithm offers improved resilience to additive noise over the PAST algorithm and it is implemented using the more stable square-root form. The asymptotic mean square error of the corresponding LPM-based model is derived and minimized at each time instant to obtain the proposed LOFF for improving the convergence speed and steady state error. A recursive bi-iteration singular value decomposition (SVD) algorithm is also proposed for recursive computation of the pseudo-inverse of the state transition matrix and its eigenvalues. This facilitates online estimation of its model order using classical model selection criteria. Moreover, a new criterion based on the percentage of explained variance is also proposed with improved performance. The SCAD regularization is proposed for automatic model selection of the input and feedthrough matrices, as it is asymptotically unbiased. Efficient techniques for incorporating SCAD and estimating other required quantities are also developed. The proposed algorithms are evaluated using computer simulations under stationary and nonstationary environments and a real dataset on wing flutter data. Results show that the proposed algorithms offer improved convergence speed and steady state performance over the conventional algorithms and it also provides an online estimate of the system model order. INDEX TERMS Recursive subspace model identification (RSI), variable forgetting factor (VFF), SCAD regularization, extended instrumental variable (EIV), local polynomial modeling (LPM), model order estimation.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: B Recursive Estimation Of a A A And C C Cmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A Variable Regularized Recursive Subspace Model Identification Algorithm With Extended Instrumental Variable and Variable Forgetting Factor

2020

View full text Add to dashboard Cite

show abstract

“…bold-italicΛn=[λK+1λK+2⋱λM×L], represents a diagonal array of M×L−K small eigenvalues. trueU^n=false[bold-italicvK+1,bold-italicvK+2,⋯,bold-italicvM×Lfalse] represents the noise subspace formed by the eigenvectors corresponding to the M×L−K eigenvalues [23]. Since the GNSS uses the spread spectrum technology, the satellite signal is embedded in the white noise, and the signal power is about 20 dB–30 dB lower than the noise power, and the influence of satellite signals on the construction of noise subspace and interference signal subspace can be ignored.…”

Section: Stc-music Algorithmmentioning

confidence: 99%

Estimation of Interference Arrival Direction Based on a Novel Space-Time Conversion MUSIC Algorithm for GNSS Receivers

Wang

Chang

et al. 2019

Sensors

View full text Add to dashboard Cite

In the estimation of the direction of arrival (DOA) for interference signals, the direction-finding error of the multiple signal classification (MUSIC) algorithm will increase in the case of multiple interferences or when the interfering signal power is weak. In this paper, a space-time conversion MUSIC (STC-MUSIC) algorithm is proposed, and the concept of a focusing parameter is introduced to improve the performance of the DOA estimation. Meanwhile, a method of variable step size peak search is proposed to reduce the amount of calculation of the STC-MUSIC algorithm. The final simulation and experimental results show that the STC-MUSIC algorithm improves the purity of the noise subspace effectively, thus improving the precision and robustness of the DOA estimation for interference signals significantly. In comparison to traditional algorithms, the convergence, stability, root mean square error (RMSE) and other performance characteristics are improved greatly.

show abstract

“…If the adjacent eigenvalues exceed one order of magnitude, the corresponding K can be used as the boundary value of eigenvalues). bold-italicUn=false[vK+1,vK+2,⋯,vMfalse] represents the noise subspace formed by the eigenvectors corresponding to the M−K eigenvalues [21,22]. According to the noise subspace bold-italicUn, the corresponding projection matrix can be obtained, bold-italicPn=bold-italicUnfalse(bold-italicUnHbold-italicUnfalse)−1bold-italicUnH…”

Section: Interference-nulling Control Algorithmmentioning

confidence: 99%

An Anti-Jamming Null-Steering Control Technique Based on Double Projection in Dynamic Scenes for GNSS Receivers

Wang

Chang

2019

Sensors

View full text Add to dashboard Cite

When a global navigation satellite system (GNSS) receiver suppresses interference in a dynamic scene, the direction of the interference signal arriving at the receiver may change rapidly. The null formed by the spatial filtering based on the array antenna will become wider and shallower, and the anti-jamming performance will deteriorate. A null-steering control technique based on a dual projection algorithm is proposed in this paper, which can effectively increase the depth of the null. In this paper, the dynamic model between the interference and the receiver is established first. Based on the model, the rate of change of the arrival direction of the interference is analyzed, and the phenomenon of the spatial filtering null becoming wider and shallower is simulated and verified. Then, the double projection algorithm is introduced to effectively deepen the null. The simulation results show that the proposed method can effectively increase the null depth by 30 to 50 dB, which significantly improves the anti-jamming performance of the spatial filtering in dynamic scenes.

show abstract

An Improved Adaptive Subspace Tracking Algorithm Based on Approximated Power Iteration

Cited by 8 publications

References 23 publications

A Variable Regularized Recursive Subspace Model Identification Algorithm With Extended Instrumental Variable and Variable Forgetting Factor

A Variable Regularized Recursive Subspace Model Identification Algorithm With Extended Instrumental Variable and Variable Forgetting Factor

Estimation of Interference Arrival Direction Based on a Novel Space-Time Conversion MUSIC Algorithm for GNSS Receivers

An Anti-Jamming Null-Steering Control Technique Based on Double Projection in Dynamic Scenes for GNSS Receivers

Contact Info

Product

Resources

About