Skip to main content
Log in

A Finite Iterative Method for Solving the General Coupled Discrete-Time Periodic Matrix Equations

  • Published:
Circuits, Systems, and Signal Processing Aims and scope Submit manuscript

Abstract

Analysis and design of linear periodic control systems are closely related to the discrete-time periodic matrix equations. In this paper, we propose an iterative algorithm based on the conjugate gradient method on the normal equations (CGNE) for finding the solution group of the general coupled periodic matrix equations

$$\begin{aligned} \left\{ \begin{array}{l} A_{1,i}X_iB_{1,i}+C_{1,i}X_{i+1}D_{1,i}=E_{1,i},\\ A_{2,i}X_iB_{2,i}+C_{2,i}X_{i+1}D_{2,i}=E_{2,i}, \end{array} \right. ~~~\mathrm {for}~~~i=1,2,3,\ldots . \end{aligned}$$

By proving some properties of the algorithm, we show that the solution group of the periodic matrix equations can be obtained within a finite number of iterations in the absence of roundoff errors. Numerical examples are given to illustrate the efficiency and accuracy of the proposed algorithm.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Fig. 1
Fig. 2

Similar content being viewed by others

References

  1. P. Benner, M.S. Hossain, Structure Preserving Iterative Solution of Periodic Projected Lyapunov Equations, in Proceedings of MATHMOD conference, Vienna 2012

  2. P. Benner, M.S. Hossain, T. Stykel, Low rank iterative methods of periodic projected Lyapunov equations and their application in model reduction of periodic descriptor systems (Chemnitz Scientific Computing Preprints, 2011)

  3. P. Benner, M.S. Hossain, T. Stykel, Model reduction of periodic descriptor systems using balanced truncation, in Model Reduction in Circuit Simulation, vol. 74, Lecture Notes in Electrical Engineering, ed. by P. Benner, M. Hinze, J. ter Maten (Springer, Berlin, 2011), pp. 193–206

    Chapter  Google Scholar 

  4. S. Bittanti, P. Colaneri, Invariant representations of discrete-time periodic systems. Automatica 36, 1777–1793 (2000)

    Article  MATH  MathSciNet  Google Scholar 

  5. R. Byers, N. Rhee, Cyclic Schur and Hessenberg Schur numerical methods for solving periodic Lyapunov and Sylvester equations. Technical Report, Department of Mathematics, University of Missouri at Kansas City, 1995

  6. E.K.W. Chu, H.Y. Fan, W.W. Lin, Projected generalized discrete-time periodic Lyapunov equations and balanced realization of periodic descriptor systems. SIAM J. Matrix Anal. Appl. 29, 982–1006 (2007)

    Article  MATH  MathSciNet  Google Scholar 

  7. E.J. Craig, The N-step iteration procedures. J. Math. Phys. 34, 64–73 (1955)

  8. M. Dehghan, M. Hajarian, Analysis of an iterative algorithm to solve the generalized coupled Sylvester matrix equations. Appl. Math. Modell. 35, 3285–3300 (2011)

    Article  MATH  MathSciNet  Google Scholar 

  9. M. Dehghan, M. Hajarian, The general coupled matrix equations over generalized bisymmetric matrices. Linear Algebra Appl. 432, 1531–1552 (2010)

    Article  MATH  MathSciNet  Google Scholar 

  10. F. Ding, T. Chen, Iterative least squares solutions of coupled Sylvester matrix equations. Syst. Control Lett. 54, 95–107 (2005)

    Article  MATH  MathSciNet  Google Scholar 

  11. F. Ding, T. Chen, On iterative solutions of general coupled matrix equations. SIAM J. Control Optim. 44, 2269–2284 (2005)

    Article  MathSciNet  Google Scholar 

  12. N.J. Fliege, Multirate Digital Signal Processing: Multirate Systems, 1st edn. (Wavelets and filter banks, New York, 1994)

    Google Scholar 

  13. R. Granat, I. Jonsson, B. Kagstrom, Recursive blocked algorithms for solving periodic triangular Sylvester-type matrix equations, in Proc. of the 8th international conference on Applied parallel computing: state of the art in scientific computing, 2006, pp. 531–539

  14. R. Granat, B. Kagstrom, Direct eigenvalue reordering in a product of matrices in periodic schur form. SIAM J. Matrix Anal. Appl. 28, 285–300 (2006)

    Article  MATH  MathSciNet  Google Scholar 

  15. R. Granat, B. Kagstrom, D. Kressner, Reordering the Eigenvalues of a Periodic Matrix Pair with Applications in Control, in Proc. of 2006 IEEE Conference on Computer Aided Control Systems Design (CACSD), 2006, pp. 25–30

  16. R. Granat, B. Kagstrom, D. Kressner, Computing periodic deflating subspaces associated with a specified set of eigenvalues. BIT Numer. Math. 43, 1–18 (2003)

    Article  Google Scholar 

  17. M.S. Hossain, P. Benner, Iterative solvers for periodic matrix equations and model reduction for periodic control systems. 7th International Conference on Electrical and Computer Engineering, Dhaka, 20–22 December, 2012

  18. D. Kressner, Large Periodic Lyapunov Equations: Algorithms and Applications, in Proc. of ECC03 (Cambridge, 2003)

  19. M. Lovera, E. De Marchi, S. Bittanti, Periodic attitude control techniques for small satellites with actuators. IEEE Trans. Control Sys.Technol. 10, 90–95 (2002)

    Article  Google Scholar 

  20. M. Pittelkau, Optimal periodic control for spacecraft pointing and attitude determination. J. Guid. Control Dyn. 16, 1078–1084 (1993)

    Article  MATH  Google Scholar 

  21. Y. Shi, F. Ding, T. Chen, \(2\)-Norm based recursive design of transmultiplexers with designable filter length. Circuits Syst. Signal Process. 25, 447–462 (2006)

    Article  MATH  Google Scholar 

  22. Y. Shi, B. Yu, Robust mixed \({\cal H_2/{\cal H}}_\infty \) control of networked control systems with random time delays in both forward and 1 backward communication links. Automatica 47, 754–760 (2011)

    Article  MATH  MathSciNet  Google Scholar 

  23. R. Smith, Matrix equation \(XA+BX =C\). SIAM J. Appl. Math. 16, 198–201 (1968)

    Article  MATH  MathSciNet  Google Scholar 

  24. J. Sreedhar, P. Van Dooren, Periodic Schur form and some matrix equations, in Proc. of the Symposium on the Mathematical Theory of Networks and Systems (MTNS’93) ed. by U. Helmke, R. Mennicken, J. Saurer, vol 1, (Regensburg, 1994), pp. 339–362

  25. T. Stykel, Low-rank iterative methods for projected generalized Lyapunov equations. Electron. Trans. Numer. Anal. 30, 187–202 (2008)

    MATH  MathSciNet  Google Scholar 

  26. A. Varga, Periodic Lyapunov equations: some applications and new algorithms. Int. J. Control 67, 69–87 (1997)

    Article  MATH  Google Scholar 

  27. A. Varga, Balancing related methods for minimal realization of periodic systems. Syst. Control Lett. 36, 339–349 (1999)

    Article  MATH  Google Scholar 

  28. A. Varga, P. Van Dooren, Computing the zeros of periodic descriptor systems. Syst. Control Lett. 50, 371–381 (2003)

    Article  MATH  Google Scholar 

  29. A. Varga, Computation of Kalman decompositions of periodic systems. Eur. J. Control 10, 1–8 (2004)

    Article  MATH  MathSciNet  Google Scholar 

  30. A.G. Wu, L.L. Lv, M.Z. Hou, Finite iterative algorithms for a common solution to a group of complex matrix equations. Appl. Math. Comput. 218, 1191–1202 (2011)

    Article  MATH  MathSciNet  Google Scholar 

  31. B. Zhou, G.R. Duan, Z.Y. Li, Gradient based iterative algorithm for solving coupled matrix equations. Syst. Control Lett. 58, 327–333 (2009)

    Article  MATH  MathSciNet  Google Scholar 

  32. B. Zhou, Z.Y. Li, G.R. Duan, Y. Wang, Weighted least squares solutions to general coupled Sylvester matrix equations. J. Comput. Appl. Math. 224, 759–776 (2009)

    Article  MATH  MathSciNet  Google Scholar 

  33. B. Zhou, Z. Lin, G.R. Duan, A parametric Lyapunov equation approach to low gain feedback design for discrete-time systems. Automatica 45, 238–244 (2009)

    Article  MATH  MathSciNet  Google Scholar 

  34. B. Zhou, G.R. Duan, Z. Lin, A parametric periodic Lyapunov equation with application in semi-global stabilization of discrete-time periodic systems subject to actuator saturation. Automatica 47, 316–325 (2011)

    Article  MATH  MathSciNet  Google Scholar 

  35. H. Zhang, Y. Shi, M.X. Liu, \({\cal H}_\infty \) step tracking control for networked discrete-time nonlinear systems with integral and predictive actions. IEEE Trans. Ind. Inform. 9, 337–345 (2013)

    Article  Google Scholar 

Download references

Acknowledgments

The author is deeply indebted to two anonymous reviewers for providing very useful comments and constructive suggestions, which greatly improved the original manuscript of this paper.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Masoud Hajarian.

Appendix

Appendix

1.1 The proof of Lemma 1

We use induction to show (2.2). For \(k=1\), we can write

$$\begin{aligned}&\sum ^\theta _{i=1}\langle S_i(1), X_i^*-X_i(1)\rangle = \sum ^\theta _{i=1}\Big \langle A_{1,i}^HR_{1,i}(1)B_{1,i}^H, X_i^*-X_i(1)\Big \rangle \nonumber \\&\qquad + \sum ^\theta _{i=1}\Big \langle C_{1,i-1}^HR_{1,i-1}(1)D_{1,i-1}^H, X_i^*-X_i(1)\Big \rangle \nonumber \\&\qquad +\sum ^\theta _{i=1}\Big \langle A_{2,i}^HR_{2,i}(1)B_{2,i}^H, X_i^*-X_i(1)\Big \rangle \nonumber \\&\qquad + \sum ^\theta _{i=1}\Big \langle C_{2,i-1}^HR_{2,i-1}(1)D_{2,i-1}^H, X_i^*-X_i(1)\Big \rangle \nonumber \\&\quad = \sum ^\theta _{i=1}\Big \langle R_{1,i}(1), A_{1,i}(X_i^*-X_i(1))B_{1,i}\Big \rangle \nonumber \\&\qquad + \sum ^\theta _{i=1}\langle R_{1,i}(1), C_{1,i}(X_{i+1}^*-X_{i+1}(1))D_{1,i}\rangle \nonumber \\&\qquad + \sum ^\theta _{i=1}\Big \langle R_{2,i}(1), A_{2,i}( X_i^*-X_i(1))B_{2,i}\Big \rangle \nonumber \\&\qquad + \sum ^\theta _{i=1}\Big \langle R_{2,i}(1), C_{2,i}(X_{i+1}^*-X_{i+1}(1))D_{2,i}\Big \rangle \nonumber \\&\quad = \sum ^\theta _{i=1}\Big \langle R_{1,i}(1), A_{1,i} X_i^*B_{1,i} +C_{1,i}X_{i+1}^*D_{1,i}\nonumber \\&\qquad - A_{1,i} X_i(1) B_{1,i} -C_{1,i}X_{i+1}(1) D_{1,i}\Big \rangle \nonumber \\&\qquad + \sum ^\theta _{i=1}\Big \langle R_{2,i}(1), A_{2,i} X_i^*B_{2,i} +C_{2,i}X_{i+1}^*D_{2,i}\nonumber \\&\qquad - A_{2,i} X_i(1) B_{2,i} -C_{2,i}X_{i+1}(1) D_{2,i} \Big \rangle \nonumber \\&\quad = \sum ^\theta _{i=1}\Big \langle R_{1,i}(1), E_{1,i} - A_{1,i} X_i(1) B_{1,i} -C_{1,i}X_{i+1}(1) D_{1,i} \Big \rangle \nonumber \\&\qquad + \sum ^\theta _{i=1}\Big \langle R_{2,i}(1), E_{2,i} - A_{2,i} X_i(1) B_{2,i} -C_{2,i}X_{i+1}(1) D_{2,i} \Big \rangle \nonumber \\&\quad = \sum ^\theta _{i=1}\left[ \Big \langle R_{1,i}(1), R_{1,i}(1) \rangle + \langle R_{2,i}(1), R_{2,i}(1) \Big \rangle \right] \nonumber \\&\quad =\sum ^\theta _{i=1}\left[ ||R_{1,i}(1)||^2+||R_{2,i}(1)||^2\right] . \end{aligned}$$
(5.1)

Suppose that the conclusion (2.2) holds for \(k=s\). When \(k=s+1\), we obtain

$$\begin{aligned}&\sum ^\theta _{i=1}\Big \langle S_i(s+1), X_i^*-X_i(s+1)\Big \rangle =\sum ^\theta _{i=1}\Big \langle A_{1,i}^HR_{1,i}(s+1)B_{1,i}^H, X_i^*-X_i(s+1)\Big \rangle \nonumber \\&\qquad +\sum ^\theta _{i=1}\Big \langle C_{1,i-1}^HR_{1,i-1}(s+1)D_{1,i-1}^H, X_i^*-X_i(s+1)\Big \rangle \nonumber \\&\qquad +\sum ^\theta _{i=1}\Big \langle A_{2,i}^HR_{2,i}(s+1)B_{2,i}^H, X_i^*-X_i(s+1)\Big \rangle \nonumber \\&\qquad +\sum ^\theta _{i=1}\Big \langle C_{2,i-1}^HR_{2,i-1}(s+1)D_{2,i-1}^H, X_i^*-X_i(s+1)\Big \rangle \nonumber \\&\qquad +\frac{\sum ^\theta _{j=1}\left[ ||R_{1,j}(s+1)||^2+||R_{2,j}(s+1)||^2\right] }{\sum ^\theta _{j=1}\left[ ||R_{1,j}(s)||^2+||R_{2,j}(s)||^2\right] }\nonumber \\&\qquad \times \sum ^\theta _{i=1}\Big \langle S_i(s), X_i^*-X_i(s+1)\Big \rangle \nonumber \\&\quad =\sum ^\theta _{i=1}\Big \langle R_{1,i}(s+1), A_{1,i} X_i^*B_{1,i}\nonumber \\&\qquad +\,C_{1,i}X_{i+1}^*D_{1,i} - A_{1,i} X_i(s+1) B_{1,i}\nonumber \\&\qquad -\,C_{1,i}X_{i+1}(s+1) D_{1,i} \Big \rangle \nonumber \\&\qquad +\, \sum ^\theta _{i=1}\Big \langle R_{2,i}(s+1), A_{2,i} X_i^*B_{2,i} +C_{2,i}X_{i+1}^*D_{2,i}\nonumber \\&\qquad - \,A_{2,i} X_i(s+1) B_{2,i}-\,C_{2,i}X_{i+1}(s+1) D_{2,i} \Big \rangle \nonumber \\&\qquad +\,\frac{\sum ^\theta _{j=1}\left[ ||R_{1,j}(s+1)||^2+||R_{2,j}(s+1)||^2\right] }{\sum ^\theta _{j=1}\left[ ||R_{1,j}(s)||^2+||R_{2,j}(s)||^2\right] }\nonumber \\&\qquad \times \sum ^\theta _{i=1}\Big \langle S_i(s), X_i^*-X_i(s)\nonumber \\&\qquad -\frac{\sum ^\theta _{j=1}\left[ ||R_{1,j}(s)||^2+||R_{2,j}(s)||^2\right] }{\sum ^\theta _{j=1}||S_j(s)||^2}S_i(s)\Big \rangle \nonumber \\&\quad =\sum ^\theta _{i=1}\left[ ||R_{1,i}(s+1)||^2+||R_{2,i}(s+1)||^2\right] \nonumber \\&\qquad +\frac{\sum ^\theta _{j=1}\left[ ||R_{1,j}(s+1)||^2+||R_{2,j}(s+1)||^2\right] }{\sum ^\theta _{j=1}\left[ ||R_{1,j}(s)||^2+||R_{2,j}(s)||^2\right] }\nonumber \\&\qquad \times \sum ^\theta _{i=1}\langle S_i(s), X_i^*-X_i(s)\rangle \nonumber \\&\qquad -\frac{\sum ^\theta _{j=1}\left[ ||R_{1,j}(s+1)||^2+||R_{2,j}(s+1)||^2\right] }{\sum ^\theta _{j=1}||S_j(s)||^2}\sum ^\theta _{i=1}\langle S_i(s), S_i(s)\rangle \nonumber \\&\quad =\sum ^\theta _{i=1}\left[ ||R_{1,i}(s+1)||^2+||R_{2,i}(s+1)||^2\right] . \end{aligned}$$
(5.2)

Therefore, the conclusion (2.2) holds by the principle of induction.

1.2 The proof of Lemma 2

Step I Noting that the inner product is commutative, it is enough to prove Lemma 2 for \(1\le u< v\le r.\) For \(u=1\) and \(v=2\), we can write

$$\begin{aligned}&\sum ^\theta _{i=1}\left[ \Big \langle R_{1,i}(1), R_{1,i}(2)\Big \rangle +\Big \langle R_{2,i}(1), R_{2,i}(2)\Big \rangle \right] \nonumber \\&\quad =\sum ^\theta _{i=1}\langle R_{1,i}(1), R_{1,i}(1)-\frac{\sum ^\theta _{j=1}\left[ ||R_{1,j}(1)||^2+||R_{2,j}(1)||^2\right] }{\sum ^\theta _{j=1}||S_j(1)||^2}\nonumber \\&\qquad \times \left[ A_{1,i}S_i(1)B_{1,i}+C_{1,i}S_{i+1}(1)D_{1,i}\right] \rangle \nonumber \\&\qquad +\sum ^\theta _{i=1}\langle R_{2,i}(1), R_{2,i}(1)-\frac{\sum ^\theta _{j=1}\left[ ||R_{1,j}(1)||^2+||R_{2,j}(1)||^2\right] }{\sum ^\theta _{j=1}||S_j(1)||^2}\nonumber \\&\qquad \times \left[ A_{2,i}S_i(1)B_{2,i}+C_{2,i}S_{i+1}(1)D_{2,i}\right] \rangle \nonumber \\&\quad =\sum ^\theta _{i=1}\left[ ||R_{1,i}(1)||+||R_{2,i}(1)||\right] -\frac{\sum ^\theta _{j=1}\left[ ||R_{1,j}(1)||^2+||R_{2,j}(1)||^2\right] }{\sum ^\theta _{j=1}||S_j(1)||^2}\nonumber \\&\qquad \times \Bigg [ \sum ^\theta _{i=1}\langle R_{1,i}(1),A_{1,i}S_i(1)B_{1,i}+C_{1,i}S_{i+1}(1)D_{1,i}\rangle \nonumber \\&\qquad +\sum ^\theta _{i=1}\langle R_{2,i}(1), A_{2,i}S_i(1)B_{2,i}+C_{2,i}S_{i+1}(1)D_{2,i}\rangle \Bigg ]\nonumber \\&\quad =\sum ^\theta _{i=1}\left[ ||R_{1,i}(1)||+||R_{2,i}(1)||\right] -\frac{\sum ^\theta _{j=1}\left[ ||R_{1,j}(1)||^2+||R_{2,j}(1)||^2\right] }{\sum ^\theta _{j=1}||S_j(1)||^2}\nonumber \\&\qquad \times \Bigg [ \sum ^\theta _{i=1}\langle A_{1,i}^HR_{1,i}(1)B_{1,i}^H,S_i(1)\rangle \nonumber \\&\qquad +\sum ^\theta _{i=1}\langle C_{1,i-1}^HR_{1,i-1}(1)D_{1,i}^H,S_{i}(1)\rangle +\sum ^\theta _{i=1}\langle A_{2,i}^HR_{2,i}(1)B_{2,i}^H,S_i(1)\rangle \nonumber \\&\qquad +\sum ^\theta _{i=1}\langle C_{2,i-1}^HR_{2,i-1}(1)D_{2,i-1}^H,S_{i}(1)\rangle \Bigg ]\nonumber \\&\quad =\sum ^\theta _{i=1}\left[ ||R_{1,i}(1)||+||R_{2,i}(1)||\right] -\frac{\sum ^\theta _{j=1}\left[ ||R_{1,j}(1)||^2+||R_{2,j}(1)||^2 \right] }{\sum ^\theta _{j=1}||S_j(1)||^2}\nonumber \\&\qquad \times \Bigg [ \sum ^\theta _{i=1}\langle A_{1,i}^HR_{1,i}(1)B_{1,i}^H+C_{1,i-1}^HR_{1,i-1}(1)D_{1,i}^H+A_{2,i}^HR_{2,i}(1)B_{2,i}^H\nonumber \\&\qquad +\, C_{2,i-1}^HR_{2,i-1}(1)D_{2,i-1}^H,S_i(1)\rangle \Bigg ]=\sum ^\theta _{i=1}\left[ ||R_{1,i}(1)||+||R_{2,i}(1)||\right] \nonumber \\&\qquad -\frac{\sum ^\theta _{j=1}\left[ ||R_{1,j}(1)||^2+||R_{2,j}(1)||^2\right] }{\sum ^\theta _{j=1}||S_j(1)||^2} \sum ^\theta _{i=1}||S_i(1)||^2=0, \end{aligned}$$
(5.3)

and

$$\begin{aligned}&\!\!\!\!\!\!\!\sum ^\theta _{i=1}\Big \langle S_i(1), S_i(2)\Big \rangle =\sum ^\theta _{i=1}\Big \langle S_i(1), A_{1,i}^HR_{1,i}(2)B_{1,i}^H+C_{1,i-1}^HR_{1,i-1}(2)D_{1,i-1}^H\nonumber \\&\!\!\!\!\!\!\!\qquad +\, A_{2,i}^HR_{2,i}(2)B_{2,i}^H+C_{2,i-1}^HR_{2,i-1}(2)D_{2,i-1}^H\Big \rangle \nonumber \\&\!\!\!\!\!\!\!\qquad +\,\frac{\sum ^\theta _{j=1}\left[ ||R_{1,j}(2)||^2+||R_{2,j}(2)||^2\right] }{\sum ^\theta _{j=1}\left[ ||R_{1,j}(1)||^2+||R_{2,j}(1)||^2\right] }\sum ^\theta _{i=1}\Big \langle S_i(1),S_i(1)\Big \rangle \nonumber \\&\!\!\!\!\!\!\!\quad =\frac{\sum ^\theta _{j=1}\left[ ||R_{1,j}(2)||^2+||R_{2,j}(2)||^2\right] }{\sum ^\theta _{j=1}\left[ ||R_{1,j}(1)||^2+||R_{2,j}(1)||^2 \right] }\sum ^\theta _{i=1}|| S_i(1)||^2 \nonumber \\&\!\!\!\!\!\!\!\qquad +\,\sum ^\theta _{i=1}\Big \langle A_{1,i}S_i(1)B_{1,i}, R_{1,i}(2) \Big \rangle \nonumber \\&\!\!\!\!\!\!\!\qquad +\,\sum ^\theta _{i=1}\Big \langle C_{1,i}S_{i+1}(1)D_{1,i}, R_{1,i}(2)\Big \rangle +\sum ^\theta _{i=1}\Big \langle A_{2,i}S_i(1)B_{2,i}, R_{2,i}(2) \Big \rangle \nonumber \\&\!\!\!\!\!\!\!\qquad +\, \sum ^\theta _{i=1}\Big \langle C_{2,i}S_{i+1}(1)D_{2,i}, R_{2,i}(2) \Big \rangle \nonumber \\&\!\!\!\!\!\!\!\quad =\frac{\sum ^\theta _{j=1}\left[ ||R_{1,j}(2)||^2+||R_{2,j}(2)||^2\right] }{\sum ^\theta _{j=1}\left[ ||R_{1,j}(1)||^2+||R_{2,j}(1)||^2\right] }\sum ^\theta _{i=1}|| S_i(1)||^2\nonumber \\&\!\!\!\!\!\!\!\qquad +\, \sum ^\theta _{i=1}\Big \langle A_{1,i}S_i(1)B_{1,i}+C_{1,i}S_{i+1}(1)D_{1,i}, R_{1,i}(2)\Big \rangle \nonumber \\&\!\!\!\!\!\!\!\qquad +\, \sum ^\theta _{i=1}\Big \langle A_{2,i}S_i(1)B_{2,i}+C_{2,i}S_{i+1}(1)D_{2,i}, R_{2,i}(2)\Big \rangle \nonumber \\&\!\!\!\!\!\!\!\quad = \frac{\sum ^\theta _{j=1}\left[ ||R_{1,j}(2)||^2+||R_{2,j}(2)||^2\right] }{\sum ^\theta _{j=1}\left[ ||R_{1,j}(1)||^2+||R_{2,j}(1)||^2\right] }\sum ^\theta _{i=1}|| S_i(1)||^2\nonumber \\&\!\!\!\!\!\!\!\qquad +\,\frac{\sum ^\theta _{j=1}||S_j(1)||^2}{\sum ^\theta _{j=1}\left[ ||R_{1,j}(1)||^2+||R_{2,j}(1)||^2\right] }\Bigg [\sum ^\theta _{i=1}\Big \langle R_{1,i}(1)-R_{1,i}(2), R_{1,i}(2)\Big \rangle \nonumber \\&\!\!\!\!\!\!\!\qquad +\, \sum ^\theta _{i=1}\Big \langle R_{2,i}(1)-R_{2,i}(2), R_{2,i}(2) \Big \rangle \Bigg ]\nonumber \\&\!\!\!\!\!\!\!\quad = \frac{\sum ^\theta _{j=1}\left[ ||R_{1,j}(2)||^2+||R_{2,j}(2)||^2\right] }{\sum ^\theta _{j=1}\left[ ||R_{1,j}(1)||^2+||R_{2,j}(1)||^2 \right] }\sum ^\theta _{i=1}|| S_i(1)||^2\nonumber \\&\!\!\!\!\!\!\!\qquad -\,\frac{\sum ^\theta _{j=1}||S_j(1)||^2}{\sum ^\theta _{j=1}\left[ ||R_{1,j}(1)||^2\!+\!||R_{2,j}(1)||^2\right] }\sum ^\theta _{i=1}\left[ ||R_{1,i}(2)||^2 +||R_{2,i}(2)||^2\right] =0.\qquad \!\!\!\quad \end{aligned}$$
(5.4)

Step II  In this step, for \(u<w<r\) we assume that

$$\begin{aligned} \sum ^\theta _{i=1}\left[ \Big \langle R_{1,i}(u), R_{1,i}(w)\Big \rangle +\Big \langle R_{2,i}(u), R_{2,i}(w)\Big \rangle \right] =0,~~~~\mathrm {and}~~~~~ \sum ^\theta _{i=1}\langle S_i(u), S_i(w)\rangle =0. \end{aligned}$$

We have

$$\begin{aligned}&\sum ^\theta _{i=1}\left[ \Big \langle R_{1,i}(u), R_{1,i}(w+1)\Big \rangle +\Big \langle R_{2,i}(u), R_{2,i}(w+1)\Big \rangle \right] \nonumber \\&\quad =\sum ^\theta _{i=1}\langle R_{1,i}(u), R_{1,i}(w)-\frac{\sum ^\theta _{j=1}\left[ ||R_{1,j}(w)||^2+||R_{2,j}(w)||^2\right] }{\sum ^\theta _{j=1}||S_j(w)||^2}\nonumber \\&\qquad \times [ A_{1,i}S_i(w)B_{1,i}+C_{1,i}S_{i+1}(w)D_{1,i}]\rangle \nonumber \\&\qquad +\sum ^\theta _{i=1}\langle R_{2,i}(u), R_{2,i}(w)-\frac{\sum ^\theta _{j=1}\left[ ||R_{1,j}(w)||^2+||R_{2,j}(w)||^2\right] }{\sum ^\theta _{j=1}||S_j(w)||^2}\nonumber \\&\qquad \times \left[ A_{2,i}S_i(w)B_{2,i}+C_{2,i}S_{i+1}(w)D_{2,i}\right] \rangle \nonumber \\&\quad =-\frac{\sum ^\theta _{j=1}\left[ ||R_{1,j}(w)||^2+||R_{2,j}(w)||^2\right] }{\sum ^\theta _{j=1}||S_j(w)||^2}\left[ \sum ^\theta _{i=1}\Big \langle A_{1,i}^HR_{1,i}(u)B_{1,i}^H, S_i(w)\Big \rangle \right. \nonumber \\&\left. \qquad +\sum ^\theta _{i=1}\Big \langle C_{1,i-1}^HR_{1,i-1}(u)D_{1,i-1}^H,S_{i}(w)\Big \rangle \right. \nonumber \\&\left. \qquad + \sum ^\theta _{i=1}\Big \langle A_{2,i}^HR_{2,i}(u)B_{2,i}^H, S_i(w)\Big \rangle + \sum ^\theta _{i=1}\Big \langle C_{2,i-1}^HR_{2,i-1}(u)D_{2,i-1}^H,S_{i}(w)\Big \rangle \right] \nonumber \\&\quad =-\frac{\sum ^\theta _{j=1}\left[ ||R_{1,j}(w)||^2+||R_{2,j}(w)||^2\right] }{\sum ^\theta _{j=1}||S_j(w)||^2}\nonumber \\&\qquad \times \left[ \sum ^\theta _{i=1}\Big \langle A_{1,i}^HR_{1,i}(u)B_{1,i}^H+C_{1,i-1}^HR_{1,i-1}(u)D_{1,i-1}^H+A_{2,i}^HR_{2,i}(u)B_{2,i}^H \right. \nonumber \\&\left. \qquad +\,C_{2,i-1}^HR_{2,i-1}(u)D_{2,i-1}^H, S_i(w)\Big \rangle \right] \nonumber \\&\quad =-\frac{\sum ^\theta _{j=1}\left[ ||R_{1,j}(w)||^2+||R_{2,j}(w)||^2\right] }{\sum ^\theta _{j=1}||S_j(w)||^2}\left[ \sum ^\theta _{i=1}\Big \langle S_i(u)\right. \nonumber \\&\left. \qquad - \frac{\sum ^\theta _{j=1}\left[ ||R_{1,j}(u)||^2+||R_{2,j}(u)||^2\right] }{\sum ^\theta _{j=1}\left[ ||R_{1,j}(u-1)||^2+||R_{2,j}(u-1)||^2\right] }S_i(u-1), S_i(w)\Big \rangle \right] =0,\nonumber \\ \end{aligned}$$
(5.5)

and

$$\begin{aligned}&\sum ^\theta _{i=1}\Big \langle S_i(u), S_i(w+1)\Big \rangle = \sum ^\theta _{i=1}\Big \langle S_i(u), A_{1,i}^HR_{1,i}(w+1)B_{1,i}^H+C_{1,i-1}^HR_{1,i-1}(w+1)D_{1,i-1}^H\nonumber \\&\qquad +\, A_{2,i}^HR_{2,i}(w+1)B_{2,i}^H+C_{2,i-1}^HR_{2,i-1}(w+1)D_{2,i-1}^H\Big \rangle \nonumber \\&\qquad +\, \frac{\sum ^\theta _{j=1}\left[ ||R_{1,j}(w+1)||^2+||R_{2,j}(w+1)||^2\right] }{\sum ^\theta _{j=1}\left[ ||R_{1,j}(w)||^2+||R_{2,j}(w)||^2\right] } \sum ^\theta _{i=1}\Big \langle S_i(u),S_i(w)\Big \rangle \nonumber \\&\quad = \sum ^\theta _{i=1}\Big \langle A_{1,i}S_i(u)B_{1,i}, R_{1,i}(w+1) \Big \rangle +\sum ^\theta _{i=1}\Big \langle C_{1,i}S_{i+1}(u)D_{1,i}, R_{1,i}(w+1)\Big \rangle \nonumber \\&\qquad +\sum ^\theta _{i=1}\Big \langle A_{2,i}S_i(u)B_{2,i}, R_{2,i}(w+1) \Big \rangle + \sum ^\theta _{i=1}\Big \langle C_{2,i}S_{i+1}(u)D_{2,i}, R_{2,i}(w+1) \Big \rangle \nonumber \\&\quad = \frac{\sum ^\theta _{j=1}||S_j(u)||^2}{\sum ^\theta _{j=1}\left[ ||R_{1,j}(u)||^2+||R_{2,j}(u)||^2\right] }\left[ \sum ^\theta _{i=1}\Big \langle R_{1,i}(u)-R_{1,i}(u+1), R_{1,i}(w+1) \Big \rangle \right. \nonumber \\&\left. \qquad + \sum ^\theta _{i=1}\Big \langle R_{2,i}(u)-R_{2,i}(u+1), R_{2,i}(w+1) \Big \rangle \right] =0. \end{aligned}$$
(5.6)

For \(u = w\), we can get

$$\begin{aligned}&\sum ^\theta _{i=1}\left[ \Big \langle R_{1,i}(w), R_{1,i}(w+1)\Big \rangle +\Big \langle R_{2,i}(w), R_{2,i}(w+1)\Big \rangle \right] \nonumber \\&\quad =\sum ^\theta _{i=1}\Big \langle R_{1,i}(w), R_{1,i}(w)-\frac{\sum ^\theta _{j=1}\left[ ||R_{1,j}(w)||^2+||R_{2,j}(w)||^2\right] }{\sum ^\theta _{j=1}||S_j(w)||^2}\nonumber \\&\qquad \times \left[ A_{1,i}S_i(w)B_{1,i}+C_{1,i}S_{i+1}(w)D_{1,i}\right] \Big \rangle \nonumber \\&\qquad +\sum ^\theta _{i=1}\Big \langle R_{2,i}(w), R_{2,i}(w)-\frac{\sum ^\theta _{j=1}\left[ ||R_{1,j}(w)||^2+||R_{2,j}(w)||^2\right] }{\sum ^\theta _{j=1}||S_j(w)||^2}\nonumber \\&\qquad \times \,[ A_{2,i}S_i(w)B_{2,i}+C_{2,i}S_{i+1}(w)D_{2,i}]\Big \rangle \nonumber \\&\quad = \sum ^\theta _{i=1}\left[ ||R_{1,i}(w)||^2+||R_{2,i}(w)||^2\right] -\frac{\sum ^\theta _{j=1}\left[ ||R_{1,j}(w)||^2+||R_{2,j}(w)||^2\right] }{\sum ^\theta _{j=1}||S_j(w)||^2}\nonumber \\&\qquad \times \left[ \sum ^\theta _{i=1}\Big \langle A_{1,i}^HR_{1,i}(w)B_{1,i}^H+C_{1,i-1}^HR_{1,i-1}(w)D_{1,i-1}^H\right. \nonumber \\&\left. \qquad +A_{2,i}^HR_{2,i}(w)B_{2,i}^H+C_{2,i-1}^HR_{2,i-1}(w)D_{2,i-1}^H, S_i(w)\Big \rangle \right] \nonumber \\&\quad = \sum ^\theta _{i=1}\left[ ||R_{1,i}(w)||^2+||R_{2,i}(w)||^2\right] \nonumber \\&\qquad -\frac{\sum ^\theta _{j=1}\left[ ||R_{1,j}(w)||^2+||R_{2,j}(w)||^2\right] }{\sum ^\theta _{j=1}||S_j(w)||^2} \sum ^\theta _{i=1}\Big \langle S_i(w)\nonumber \\&\qquad -\frac{\sum ^\theta _{j=1}\left[ ||R_{1,j}(w)||^2+||R_{2,j}(w)||^2\right] }{\sum ^\theta _{j=1}\left[ ||R_{1,j}(w-1)||^2+||R_{2,j}(w-1)||^2\right] }S_i(w-1), S_i(w)\Big \rangle \nonumber \\&\quad =\sum ^\theta _{i=1}\left[ ||R_{1,i}(w)||^2+||R_{2,i}(w)||^2\right] \nonumber \\&\qquad -\frac{\sum ^\theta _{j=1}\left[ ||R_{1,j}(w)||^2+||R_{2,j}(w)||^2\right] }{\sum ^\theta _{j=1}||S_j(w)||^2} \sum ^\theta _{i=1}||S_i(w)||^2=0, \end{aligned}$$
(5.7)
$$\begin{aligned}&\sum ^\theta _{i=1}\Big \langle S_i(w), S_i(w+1)\Big \rangle \nonumber \\&\quad = \sum ^\theta _{i=1}\Big \langle S_i(w), A_{1,i}^HR_{1,i}(w+1)B_{1,i}^H+C_{1,i-1}^HR_{1,i-1}(w+1)D_{1,i-1}^H\nonumber \\&\qquad + A_{2,i}^HR_{2,i}(w+1)B_{2,i}^H+C_{2,i-1}^HR_{2,i-1}(w+1)D_{2,i-1}^H\Big \rangle \nonumber \\&\qquad + \frac{\sum ^\theta _{j=1}\left[ ||R_{1,j}(w+1)||^2+||R_{2,j}(w+1)||^2\right] }{\sum ^\theta _{j=1}\left[ ||R_{1,j}(w)||^2+||R_{2,j}(w)||^2\right] } \sum ^\theta _{i=1}\Big \langle S_i(w),S_i(w)\Big \rangle \nonumber \\&\quad = \frac{\sum ^\theta _{j=1}\left[ ||R_{1,j}(w+1)||^2+||R_{2,j}(w+1)||^2\right] }{\sum ^\theta _{j=1}\left[ ||R_{1,j}(w)||^2+||R_{2,j}(w)||^2\right] } \sum ^\theta _{i=1}||S_i(w)||^2\nonumber \\&\qquad + \sum ^\theta _{i=1}\Big \langle A_{1,i}S_i(w)B_{1,i}, R_{1,i}(w+1) \Big \rangle +\sum ^\theta _{i=1}\Big \langle C_{1,i}S_{i+1}(w)D_{1,i}, R_{1,i}(w+1)\Big \rangle \nonumber \\&\qquad +\sum ^\theta _{i=1}\Big \langle A_{2,i}S_i(w)B_{2,i}, R_{2,i}(w+1) \Big \rangle + \sum ^\theta _{i=1}\Big \langle C_{2,i}S_{i+1}(w)D_{2,i}, R_{2,i}(w+1) \Big \rangle \nonumber \\&\quad = \frac{\sum ^\theta _{j=1}\left[ ||R_{1,j}(w+1)||^2+||R_{2,j}(w+1)||^2\right] }{\sum ^\theta _{j=1}\left[ ||R_{1,j}(w)||^2+||R_{2,j}(w)||^2\right] } \sum ^\theta _{i=1}||S_i(w)||^2\nonumber \\&\qquad + \frac{\sum ^\theta _{j=1}||S_j(w)||^2}{\sum ^\theta _{j=1}\left[ ||R_{1,j}(w)||^2+||R_{2,j}(w)||^2\right] }\left[ \sum ^\theta _{i=1}\Big \langle R_{1,i}(w)-R_{1,i}(w+1), R_{1,i}(w+1) \Big \rangle \right. \nonumber \\&\left. \qquad + \sum ^\theta _{i=1}\Big \langle R_{2,i}(w)-R_{2,i}(w+1), R_{2,i}(w+1) \Big \rangle \right] \nonumber \\&\quad = \frac{\sum ^\theta _{j=1}\left[ ||R_{1,j}(w+1)||^2+||R_{2,j}(w+1)||^2\right] }{\sum ^\theta _{j=1}\left[ ||R_{1,j}(w)||^2+||R_{2,j}(w)||^2\right] } \sum ^\theta _{i=1}||S_i(w)||^2\nonumber \\&\qquad - \frac{\sum ^\theta _{j=1}||S_j(w)||^2}{\sum ^\theta _{j=1}\left[ ||R_{1,j}(w)||^2+||R_{2,j}(w)||^2\right] }\sum ^\theta _{i=1}\left[ ||R_{1,i}(w+1)||^2+||R_{2,i}(w+1)||^2\right] =0. \nonumber \\ \end{aligned}$$
(5.8)

By considering Steps I and II, two statements of Lemma 2 hold by the principle of induction.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Hajarian, M. A Finite Iterative Method for Solving the General Coupled Discrete-Time Periodic Matrix Equations. Circuits Syst Signal Process 34, 105–125 (2015). https://doi.org/10.1007/s00034-014-9842-1

Download citation

  • Received:

  • Revised:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s00034-014-9842-1

Keywords

Navigation