Table of Contents Author Guidelines Submit a Manuscript
Abstract and Applied Analysis
Volume 2015 (2015), Article ID 471362, 8 pages
http://dx.doi.org/10.1155/2015/471362
Research Article

Analysis of the Structured Perturbation for the BCSCB Linear System

1Department of Mathematics, Linyi University, Linyi, Shandong 276000, China
2Department of Mathematics, Shandong Normal University, Jinan, Shandong 250014, China

Received 22 July 2014; Accepted 10 September 2014

Academic Editor: Zidong Wang

Copyright © 2015 Xia Tang and Zhaolin Jiang. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

Circulant and block circulant type matrices are important tools in solving networked systems. In this paper, based on the style spectral decomposition of the basic circulant matrix and the basic skew circulant matrix, the block style spectral decomposition of the BCSCB matrix is obtained. And then, the structure perturbation is analysed, which includes the condition number and relative error of the BCSCB linear system. Then the optimal backward perturbation bound of the BCSCB linear system is discussed. Simultaneously, the algorithm for the optimal backward perturbation bound is given. Finally, a numerical example is provided to verify the effectiveness of the algorithm.

1. Introduction

It is an active objective that circulant and block circulant type matrices are applied to networks engineering. The stability region in the parameters space is extended by the breaking of a delayed ring neural network where the form of time-delay systems is , where is a circulant matrix, if the number of the neurons is sufficiently large in [1]. In [2], the question of when circulant quantum spin networks with nearest-neighbor couplings can give perfect state transfer is solved. The properties of linear diffusion algorithm are investigated both by a worst-case analysis and by a probabilistic analysis and are shown to depend on the spectral properties of the circulant matrix in [3]. A viable option for increasing the lifetime of the sensor network for a small loss in accuracy of the query results whose matrices are circulant is offered in [4]. In [5], the authors considered the kinetics of an autocatalytic reaction network in which replication and catalytic actions are separated by a translation step. They found that the behavior of such a system is closely related to second-order replicator equations, where the second-order replicator equations are circulant interaction matrices. In order to obtain the optimal routing in double loop networks, the problem of finding the shortest path in circulant graphs with an arbitrary number of jumps is studied in [6].

A block circulant with skew circulant blocks matrix with the first row has the following form:and for any , The matrix is denoted by BCSCB.

Rigal and Gaches [7] considered a posteriori analysis of the compatibility of a computed solution to the uncertain data of a linear system by some new theorems generalizing a result of Oettli and Prager. In [8], the style spectral decomposition of the skew circulant matrix is given and the optimal backward perturbation analysis for the skew circulant linear system is discussed. Liu and Guo [9] obtained the bound of the optimal backward perturbation for a block circulant linear system. J.-G. Sun and Z. Sun [10] studied the optimal backward perturbation bounds for undetermined systems. In [11], the optimal backward perturbation analysis for the block skew circulant linear system with skew circulant blocks is given by Li et al.

2. The Block Style Spectral Decomposition of the BCSCB Matrix

Let matrix be a BCSCB matrix as the form (1); then by using the properties of Kronecker products in [12], the can be decomposed as where is a square matrix of order , and it has the following form:

Based on and in [9], the style spectral decomposition of the matrix is where is an orthogonal matrix.

When is even,

When is odd,

Taking (3) and (5) into consideration, the matrix can be decomposed as is an orthogonal matrix obviously. So (8) is the block style spectral decomposition of the matrix .

3. Analysis of the Structured Perturbation

The structured perturbation analysis for BCSCB linear system is given in this section. We discuss the condition number and the relative error of the BCSCB linear system. The optimal backward perturbation bound of the BCSCB linear system is analysed. And, at the end of the section, we give the algorithm for the optimal backward perturbation bound.

3.1. Condition Number and Relative Error of BCSCB Linear System

Consider where is defined in (1).

From (8) and the property of Kronecker products in [12], the matrix can be expressed by using the elements in its first row as where is a square matrix of order , and it has the following form:

Based on and in [8], the style spectral decomposition of the matrix is where is an orthogonal matrix.

When is even,

When is odd,

Furthermore, (10) can be expressed as follows: and here , where and are identity matrices with orders and , respectively.

The problem will be discussed at two different situations.

(1) When is even,

(2) When is odd,

We denote by the eigenvalues of matrix [9], and are denoted as the eigenvalues of matrix [8], and then the eigenvalues of are obtained (refer to [12, 13]). Consider

Lemma 1. is a nonsingular matrix if and only if , where

Let

Theorem 2. If , then the singular values of the matrix are .

Proof. Assume the conjugate transpose of is By a direct calculation, is a normal matrix as . Then matrix is a unitarily diagonalizable matrix based on Theorem in [14]. Then there exists a unitary matrix , such that where are the eigenvalues of matrix . Taking the conjugate transpose at both sides of (22), we get then And are the eigenvalues of the matrix , for any . Therefore, the singular values of are Recall (19) and (20); the proof is completed.

As the definition of the spectral norm of matrix is via Theorem 2, the following corollary is obtained.

Corollary 3. Let ; then the spectrum norm of matrix is

Let be a perturbation of the coefficient matrix and let be a perturbation of the vector , where has the following form: and for any , Let If then through Lemma 1, is a nonsingular matrix. Let By , we obtain where is a BCSCB matrix apparently, and . So The following theorem can be obtained.

Theorem 4. Let , , , , and be defined as above. If , then where

Remark 5. The condition number of the BCSCB matrix can be easily computed with the basis of (37) and (38), the same as the bound of perturbation (37).

3.2. Optimal Backward Perturbation Bound of the BCSCB Linear System

In this part, a new method is given to obtain the minimal value of the perturbation bound, which is only related to the perturbation of the coefficient matrix and the vector. At the end of this part, the algorithm for the optimal backward perturbation bound is given.

Let be an approximate solution to and let which is equal to According to [7], we can get

Let be an approximate solution to , where is defined in (1): So (as is a BCSCB matrix, ). Hence, Since the question will be analysed in two different conditions.

(1) When is even, where

(2) When is odd, where

Let and then which is equal to As is a convex function of , the point of the minimal value is Substituting it back into (49), we obtain the following.

Theorem 6. Consider

Let be the singular value decomposition of , where and are unitary matrices, , and ; then where , , .

Remark 7. As , then can be obtained; hence, .

From what we analysed above, the following algorithm can be obtained.

Algorithm 8. We have the following steps.
Step 1. Form the style spectral decomposition of the matrixes and :Step 2. Form the block style spectral decomposition of the BCSCB matrix.
Step 3. Compute .
Step 4. Compute .
Step 5. Compute Step 6. Form .
Step 7. Compute the singular value decomposition of .
Step 8. Compute .

4. Numerical Example

In this section, a simple numerical example is given to verify the conclusion above. Suppose that , in the following example.

If the coefficient matrix of the BCSCB linear system is and the constant vector , now, three perturbations are given as follows: From the equation , where , are defined as above, the approximate solution of can be obtained as follows:where is the solution of and , , is the solution of , , respectively.

Based on Algorithm 8, we obtain Table 1, where is the relative error of the BCSCB linear system, is the condition number, , and can be obtained from the algorithm.

Table 1: The related date of the algorithm.

From the numerical example, the accuracy of the conclusion and the effectiveness of the algorithm are verified.

5. Conclusion

In this paper, we consider the problems associated with the BCSCB matrix. The BCSCB matrix is an extension of the circulant matrix and skew circulant matrix. We give the form of the BCSCB matrix and obtain its block style spectral decomposition. The algorithm of the optimal backward perturbation is given. Furthermore, by circulant matrices technology, we will develop solving problems in [1517].

Conflict of Interests

The authors declare that there is no conflict of interests regarding the publication of this paper.

Acknowledgments

This research was supported by the Development Project of Science & Technology of Shandong Province (Grant no. 2012GGX10115) and the AMEP of Linyi University, China.

References

  1. T. N. Khokhlova and M. M. Kipnis, “The breaking of a delayed ring neural network contributes to stability: the rule and exceptions,” Neural Networks, vol. 48, pp. 148–152, 2013. View at Publisher · View at Google Scholar · View at Scopus
  2. M. Bašić, “Characterization of quantum circulant networks having perfect state transfer,” Quantum Information Processing, vol. 12, no. 1, pp. 345–364, 2013. View at Publisher · View at Google Scholar · View at MathSciNet · View at Scopus
  3. P. Frasca, R. Carli, F. Fagnani, and S. Zampieri, “Average consensus on networks with quantized communication,” International Journal of Robust and Nonlinear Control, vol. 19, no. 16, pp. 1787–1816, 2009. View at Publisher · View at Google Scholar · View at MathSciNet · View at Scopus
  4. V. Puttagunta and K. Kalpakis, “Accuracy vs. lifetime: linear sketches for aggregate queries in sensor networks,” Algorithmica, vol. 49, no. 4, pp. 357–385, 2007. View at Publisher · View at Google Scholar · View at MathSciNet · View at Scopus
  5. R. Happel, R. Hecht, and P. F. Stadler, “Autocatalytic networks with translation,” Bulletin of Mathematical Biology, vol. 58, no. 5, pp. 877–905, 1996. View at Publisher · View at Google Scholar · View at Zentralblatt MATH · View at Scopus
  6. D. Gómez, J. Gutierrez, and Á. Ibeas, “Optimal routing in double loop networks,” Theoretical Computer Science, vol. 381, no. 1–3, pp. 68–85, 2007. View at Publisher · View at Google Scholar · View at MathSciNet · View at Scopus
  7. J. L. Rigal and J. Gaches, “On the compatibility of a given solution with the data of a linear system,” Journal of the Association for Computing Machinery, vol. 14, pp. 543–548, 1967. View at Publisher · View at Google Scholar · View at MathSciNet
  8. J. Li, Z. Jiang, N. Shen, and J. Zhou, “On optimal backward perturbation analysis for the linear system with skew circulant coefficient matrix,” Computational and Mathematical Methods in Medicine, vol. 2013, Article ID 707381, 7 pages, 2013. View at Publisher · View at Google Scholar · View at MathSciNet
  9. X. Liu and X. Guo, “On optimal backward perturbation analysis for the linear system with block cyclic coefficient matrix,” Numerical Mathematics, vol. 12, no. 2, pp. 162–172, 2003. View at Google Scholar · View at MathSciNet
  10. J.-G. Sun and Z. Sun, “Optimal backward perturbation bounds for underdetermined systems,” SIAM Journal on Matrix Analysis and Applications, vol. 18, no. 2, pp. 393–402, 1997. View at Publisher · View at Google Scholar · View at MathSciNet · View at Scopus
  11. Z. Jiang, J. Li, and J. Zhou, “Optimal backward perturbation analysis for the block skew circulant linear systems with skew circulant blocks,” Abstract and Applied Analysis, vol. 2014, Article ID 523102, 8 pages, 2014. View at Publisher · View at Google Scholar · View at MathSciNet
  12. Z. L. Jiang and Z. X. Zhou, Circulant Matrix, Chengdu Technology University, Chengdu, China, 1999.
  13. Z. L. Jiang and Z. X. Zhou, “Nonsingularity on level-2(r1, r2)-circulant matrices of type (m,n),” Chinese Quarterly Journal of Mathematics, vol. 11, no. 2, pp. 106–110, 1996. View at Google Scholar
  14. R. A. Horn and C. R. Johnson, Matrix Analysis, Posts and Telecom Press, 2005.
  15. H. Dong, Z. Wang, and H. Gao, “Distributed H filtering for a class of markovian jump nonlinear time-delay systems over lossy sensor networks,” IEEE Transactions on Industrial Electronics, vol. 60, no. 10, pp. 4665–4672, 2013. View at Publisher · View at Google Scholar · View at Scopus
  16. Z. Wang, H. Dong, B. Shen, and H. Gao, “Finite-horizon H\sb filtering with missing measurements and quantization effects,” IEEE Transactions on Automatic Control, vol. 58, no. 7, pp. 1707–1718, 2013. View at Publisher · View at Google Scholar · View at MathSciNet · View at Scopus
  17. D. Ding, Z. Wang, J. Hu, and H. Shu, “Dissipative control for state-saturated discrete time-varying systems with randomly occurring nonlinearities and missing measurements,” International Journal of Control, vol. 86, no. 4, pp. 674–688, 2013. View at Publisher · View at Google Scholar · View at MathSciNet · View at Scopus