Structured Primal-dual Interior-point Methods for Banded Semidefinite Programming

  • Zhiming Deng
  • Ming Gu
  • Michael L. Overton
Part of the Operator Theory: Advances and Applications book series (OT, volume 202)


For semidefinite programming (SDP) problems, traditional primal-dual interior-point methods based on conventional matrix operations have an upper limit on the problem size that the computer can handle due to memory constraints. But for a special kind of SDP problem, which is called the banded symmetric semidefinite programming (BSDP) problem, a memory-efficient algorithm, called a structured primal-dual interior-point method, can be applied. The method is based on the observation that both banded matrices and their inverses can be represented in sequentially semi-separable (SSS) form with numerical ranks equal to the half bandwidths of the banded matrices. Moreover, all computation can be done sequentially using the SSS form. Experiments of various problem sizes are performed to verify the feasibility of the proposed method.


Banded matrix semidefinite program interior-point method sequentially semi-separable 

Mathematics Subject Classification (2000)

65F05 90C22 90C51 65F99 90C25 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. [1]
    F. Alizadeh, J.-P. Haeberly and M. Overton, Primal-Dual Interior-Point Methods For Semidefinite Programming: Convergence Rates, Stability And Numerical Results, SIAM J. Optim. p. 746–768, vol. 8, No. 3, Aug 1998.zbMATHCrossRefMathSciNetGoogle Scholar
  2. [2]
    R.B. Bapat, On generalized inverse of banded matrices, Electronic J. Line. Alge. p. 284–290, vol. 16, Sep. 2007.zbMATHMathSciNetGoogle Scholar
  3. [3]
    S. Chandrasekaran, P. Dewilde, M. Gu, T. Pals, X. Sun, A.-J. van der Veen and D. White, Fast Stable Solvers for Sequentially Semi-Separable Linear Systems of Equations and Linear Squares Problems, University of California at Berkeley, 2003.Google Scholar
  4. [4]
    S. Chandrasekaran, P. Dewilde, M. Gu, T. Pals, X. Sun, A.-J. van der Veen and D. White, Some Fast Algorithms for Sequentially Semiseparable Representations, SIAM J. Matrix Anal. Appl., p. 341–364, vol. 27, No. 2, Sep. 2005.zbMATHCrossRefMathSciNetGoogle Scholar
  5. [5]
    P. Dewilde and A. van der Veen, Time-varying systems and computations, Kluwer Academic Publishers, 1998.Google Scholar
  6. [6]
    I. Gohberg, T. Kailath and I. Koltracht, Linear complexity algorithms for semiseparable matrices, Integral Equations Operator Theory, p. 780–804, vol. 8, 1985.zbMATHCrossRefMathSciNetGoogle Scholar
  7. [7]
    R.A. Horn and C.R. Johnson, Topics in Matrix Analysis, Cambridge University Press, Cambridge, 1991.zbMATHGoogle Scholar
  8. [8]
    E. de Klerk, Aspects of semidefinite programming: Interior Point Algorithms and Selected Applications, Kluwer Academic Publishers, 2002.Google Scholar
  9. [9]
    E. de Klerk, Exploiting special structure in semidefinite programming. European Journal of Operational Research, 2009. To appear.Google Scholar
  10. [10]
    C. Roos, T. Terlaky and J.-P. Vial, Interior Point Methods For Linear Optimization, Springer, 2006.Google Scholar
  11. [11]
    G. Strang and T. Nguyen, The interplay of ranks of submatrices, SIAM Review, p. 637–646, vol. 46, No. 4, 2004.zbMATHCrossRefMathSciNetGoogle Scholar
  12. [12]
    M.J. Todd, Semidefinite Optimization, Acta Numerica, p. 515–560, vol. 10, 2001.zbMATHCrossRefMathSciNetGoogle Scholar

Copyright information

© Birkhäuser Verlag Basel/Switzerland 2010

Authors and Affiliations

  • Zhiming Deng
    • 1
  • Ming Gu
    • 2
  • Michael L. Overton
    • 3
  1. 1.Berkeley Wireless Research CenterBerkeleyUSA
  2. 2.Department of MathematicsUniversity of California at BerkeleyBerkeleyUSA
  3. 3.Courant Institute of Mathematical SciencesNew York UniversityNew YorkUSA

Personalised recommendations