# Joint diagonalization via subspace fitting techniques

@inproceedings{Veen2001JointDV, title={Joint diagonalization via subspace fitting techniques}, author={Alle-Jan van der Veen}, booktitle={ICASSP}, year={2001} }

Joint diagonalization problems of Hermitian or non-Hermitian matrices occur as the final parameter estimation step in several blind source separation problems such as ACMA, JADE, PARAFAC, and SOBI. Previous approaches have been Jacobi iteration schemes and alternating projections. Here we show how the joint diagonalization problem can be formulated as a (weighted) subspace fitting problem so that it can be solved using the efficient Gauss-Newton optimization algorithm proposed for that problem… Expand

#### Figures and Topics from this paper

#### 81 Citations

A Direct Algorithm for Nonorthogonal Approximate Joint Diagonalization

- Computer Science, Mathematics
- IEEE Transactions on Signal Processing
- 2012

This work proposes a suboptimal but closed-form solution for AJD in the direct least-squares sense and gives the acronym DIEM (DIagonalization using Equivalent Matrices), which is both fast and accurate compared to the state-of-the-art iterative AJD algorithms. Expand

Joint diagonalization of correlation matrices by using Newton methods with application to blind signal separation

- Sensor Array and Multichannel Signal Processing Workshop Proceedings, 2002
- 2002

The paper addresses the blind signal separation problem in the presence of sensor noise for the case where the source signals are non-stationary and/or non-white. This problem can be formulated as a… Expand

JACOBI-LIKE NONNEGATIVE JOINT DIAGONALIZATION BY CONGRUENCE

- 2013

A new joint diagonalization by congruence algorithm is presented, which allows the computation of a nonnegative joint diagonalizer. The nonnegativity constraint is ensured by means of a square change… Expand

A New Joint Diagonalization Algorithm with Application to Blind Source Separation

- Computer Science
- 2012 8th International Conference on Wireless Communications, Networking and Mobile Computing
- 2012

A novel iterative algorithm for the joint diagonalization of a set of real symmetric matrices is presented and an explicit method for optimizing the cost function is derived and compared with other techniques in the BSS context. Expand

A New Joint Diagonalization Algorithm with Application to Blind Source Separation

- Computer Science
- 2012

A novel iterative algorithm for the joint diagonalization of a set of real symmetric matrices is presented and an explicit method for optimizing the cost function is derived and compared with other techniques in the BSS context. Expand

Nonorthogonal Joint Diagonalization Algorithm Based on Trigonometric Parameterization

- Mathematics, Computer Science
- IEEE Transactions on Signal Processing
- 2007

A new approach is presented to joint diagonalization for a set of symmetric matrices with a general (and not necessarily orthogonal) matrix via a series of asymmetric eigen decompositions, including merits of simplicity, effectiveness, and computational efficiency. Expand

Joint diagonalization of correlation matrices by using gradient methods with application to blind signal separation

- Computer Science
- Sensor Array and Multichannel Signal Processing Workshop Proceedings, 2002
- 2002

The paper addresses the blind signal separation problem for the case where the source signals are non-stationary and/or non-white, and the sensors are possibly noisy by presenting cost functions for jointly diagonalizing several correlation matrices. Expand

A new Jacobi-like nonnegative joint diagonalization by congruence

- Computer Science, Mathematics
- 21st European Signal Processing Conference (EUSIPCO 2013)
- 2013

A new joint diagonalization by congruence algorithm is presented, which allows the computation of a nonnegative joint diagonalizer and a Jacobi-like approach using LU matrix factorization, which consists of formulating a high-dimensional optimization problem into several sequential one-dimensional subproblems. Expand

Fast Approximate Joint Diagonalization of Positive Definite Hermitian Matrices

- Mathematics, Computer Science
- 2007 IEEE International Conference on Acoustics, Speech and Signal Processing - ICASSP '07
- 2007

The results imply that the SVDJD algorithm is computationally efficient with performance similar to state-of-the-art algorithms for approximate joint diagonalization. Expand

Non-orthogonal joint diagonalization in the least-squares sense with application in blind source separation

- Mathematics, Computer Science
- IEEE Trans. Signal Process.
- 2002

This work proposes an iterative alternating-directions algorithm for minimizing the WLS criterion with respect to a general (not necessarily orthogonal) diagonalizing matrix and proves weak convergence in the sense that the norm of parameters update is guaranteed to fall below any arbitrarily small threshold within a finite number of iterations. Expand

#### References

SHOWING 1-10 OF 13 REFERENCES

A least-squares approach to joint Schur decomposition

- Mathematics, Computer Science
- Proceedings of the 1998 IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP '98 (Cat. No.98CH36181)
- 1998

New Jacobi-like algorithms that extend and improve the existing JSD algorithms are proposed that are shown to coincide with that proposed intuitively by Haardt et al. (1996), thus establishing the optimality of their criterion in the least-squares sense. Expand

In-variance of subspace based estimators

- Mathematics, Computer Science
- IEEE Trans. Signal Process.
- 2000

The generalization of the optimal weighted subspace fitting approach introduced by Viberg (1989) in the DOA estimation context is discussed; some invariance properties of optimally weighted estimate are exhibited; and the equivalence between subspacefitting and subspace matching is shown. Expand

Simultaneous Schur decomposition of several nonsymmetric matrices to achieve automatic pairing in multidimensional harmonic retrieval problems

- Mathematics, Computer Science
- IEEE Trans. Signal Process.
- 1998

A new Jacobi-type method to calculate a simultaneous Schur decomposition (SSD) of several real-valued, nonsymmetric matrices by minimizing an appropriate cost function, which achieves automatic pairing of the estimated R-dimensional modes via a closed-form procedure that neither requires any search nor any other heuristic pairing strategy. Expand

A blind source separation technique using second-order statistics

- Mathematics, Computer Science
- IEEE Trans. Signal Process.
- 1997

A new source separation technique exploiting the time coherence of the source signals is introduced, which relies only on stationary second-order statistics that are based on a joint diagonalization of a set of covariance matrices. Expand

Blind beamforming for non-gaussian signals

- Mathematics
- 1993

The paper considers an application of blind identification to beamforming. The key point is to use estimates of directional vectors rather than resort to their hypothesised value. By using estimates… Expand

An analytical constant modulus algorithm

- Mathematics, Computer Science
- IEEE Trans. Signal Process.
- 1996

It is shown that the underlying constant modulus factorization problem is, in fact, a generalized eigenvalue problem, and may be solved via a simultaneous diagonalization of a set of matrices. Expand

Detection and estimation in sensor arrays using weighted subspace fitting

- Mathematics, Computer Science
- IEEE Trans. Signal Process.
- 1991

A Gauss-Newton-type method is presented for solving the multidimensional WSF and maximum-likelihood optimization problems and a scheme for consistently estimating this is proposed based on an asymptotic analysis of the WSF cost function. Expand

Joint angle and delay estimation using shift-invariance techniques

- Mathematics, Computer Science
- IEEE Trans. Signal Process.
- 1998

The algorithm uses a two-dimensional ESPRIT-like shift-invariance technique to separate and estimate the phase shifts due to delay and direction of incidence with automatic pairing of the two parameter sets. Expand

Parallel factor analysis in sensor array processing

- Mathematics, Computer Science
- IEEE Trans. Signal Process.
- 2000

This link facilitates the derivation of powerful identifiability results for MI-SAP, shows that the uniqueness of single- and multiple-invariance ESPRIT stems from uniqueness of low-rank decomposition of three-way arrays, and allows tapping on the available expertise for fitting the PARAFAC model. Expand

Azimuth and elevation computation in high resolution DOA estimation

- Computer Science
- IEEE Trans. Signal Process.
- 1992

A number of high-resolution direction finding methods for determining the two-dimensional DOA (directions of arrival) of a number of plane waves impinging on a sensor array are discussed. The array… Expand