pith. sign in

arxiv: 1907.00655 · v1 · pith:WLJMV7FXnew · submitted 2019-07-01 · 🧮 math.OC

An entropy-based bound for the computational complexity of a switched system

Pith reviewed 2026-05-25 12:16 UTC · model grok-4.3

classification 🧮 math.OC
keywords joint spectral radiusswitched systemssum of squaresp-radiusconstrained switchingLyapunov functionsstability analysisentropy of languages
0
0 comments X

The pith

The gap between the sum-of-squares upper bound on the joint spectral radius and the true value is controlled by the p-radius and the entropy of the allowed switching sequences.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper shows that for switched systems with constraints on allowed switching sequences, the popular sum-of-squares method for finding Lyapunov functions yields an upper bound on the joint spectral radius whose distance to the true value can be bounded explicitly. The bound uses two quantities: the p-radius, which measures growth under a weighted norm, and the entropy of the language of permitted switches, which captures how many sequences are allowed. This guarantee matters because it turns an otherwise heuristic approximation into one with a computable error term. The authors also give a reduction that converts the joint spectral radius problem for low-rank matrices into a constrained joint spectral radius problem on matrices of lower dimension.

Core claim

For constrained switched systems the sum-of-squares relaxation supplies an upper bound on the joint spectral radius whose accuracy is guaranteed by the p-radius of the system together with the entropy of the language of allowed switching sequences; the same framework yields a dimension-reduction procedure that converts the joint spectral radius of low-rank matrices into the constrained joint spectral radius of matrices of strictly smaller size.

What carries the argument

The p-radius combined with the entropy of the language of allowed switching sequences, which together quantify the gap between the sum-of-squares upper bound and the true joint spectral radius.

If this is right

  • The sum-of-squares method supplies a provably accurate approximation whose error shrinks when the entropy of allowed switches is small.
  • Stability certificates obtained via sum of squares can be accompanied by an explicit numerical guarantee on their conservatism.
  • The joint spectral radius of any set of low-rank matrices can be recovered from the constrained joint spectral radius of a smaller set obtained by the reduction.
  • Constrained joint spectral radius computations become a primitive that solves a wider class of unconstrained problems.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • If the entropy can be computed or bounded for common constraint languages, the method yields practical a-priori error estimates without solving the joint spectral radius exactly.
  • The reduction step suggests that rank-deficient matrix sets arising in applications can be handled by first projecting onto a lower-dimensional constrained problem.
  • The same entropy-based accounting might extend to other convex relaxations beyond sum of squares, provided an analogous radius quantity exists.

Load-bearing premise

The p-radius and the entropy of the allowed switching language are treated as known or computable quantities that can be used to quantify the gap.

What would settle it

A concrete constrained switched system for which the observed gap between the sum-of-squares upper bound and the true joint spectral radius exceeds the value predicted from its p-radius and switching-language entropy.

Figures

Figures reproduced from arXiv: 1907.00655 by Beno\^it Legat, Pablo A. Parrilo, Rapha\"el M. Jungers.

Figure 1
Figure 1. Figure 1: Automaton for the running example. The numbers [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Representation of the solutions to Program 1 with [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Result of Example 4 for d = 1, 2, 3, 4, 5, 6; the value of d is given in the horizontal axis. The exact value of the CJSR found in [30] is represented by the horizontal line. The upper and lower bounds given by Lemma 2 using the 2d-radius are denoted “2d-radius”. The upper bound found by Program 1 with polynomials of degree 2d is denoted “Upper Bound”. From this upper bound, three lower bounds can be obtai… view at source ↗
Figure 4
Figure 4. Figure 4: Simple example of the low rank reduction. [PITH_FULL_IMAGE:figures/full_fig_p007_4.png] view at source ↗
read the original abstract

The joint spectral radius (JSR) of a set of matrices characterizes the maximal asymptotic growth rate of an infinite product of matrices of the set. This quantity appears in a number of applications including the stability of switched and hybrid systems. A popular method used for the stability analysis of these systems searches for a Lyapunov function with convex optimization tools. We analyse the accuracy of this method for constrained switched systems, a class of systems that has attracted increasing attention recently. We provide a new guarantee for the upper bound provided by the sum of squares implementation of the method. This guarantee relies on the p-radius of the system and the entropy of the language of allowed switching sequences. We end this paper with a method to reduce the computation of the JSR of low rank matrices to the computation of the constrained JSR of matrices of small dimension.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. The manuscript analyzes the accuracy of sum-of-squares (SOS) Lyapunov relaxations for upper-bounding the constrained joint spectral radius (JSR) of switched systems. It claims a new a-priori guarantee on the gap between the SOS upper bound and the true constrained JSR, expressed in terms of the p-radius of the matrix set and the topological entropy of the admissible switching language. The paper also presents a reduction showing that the (unconstrained) JSR of a set of low-rank matrices can be recovered from the constrained JSR of a related set of matrices of strictly smaller dimension.

Significance. If the central guarantee is correct and the p-radius together with the switching entropy can be obtained independently, the result supplies a concrete, non-asymptotic error bound for a widely used convex relaxation; this would be a useful theoretical tool for stability certification of constrained hybrid systems. The reduction for low-rank matrices is a self-contained algorithmic contribution that may lower the effective dimension of certain JSR instances. Both contributions are stated at a level of generality that could apply beyond the SOS setting.

major comments (2)
  1. [Abstract] Abstract, paragraph 3: the claimed guarantee is expressed solely in terms of the p-radius and the entropy of the admissible language, yet the manuscript supplies neither an algorithm nor a reduction showing that either quantity is computable by a procedure whose complexity is strictly lower than that of the original constrained-JSR problem; without such a demonstration the bound remains non-constructive and does not tighten the practical analysis of SOS relaxations.
  2. [Final section] The reduction for low-rank matrices (final section) is stated without an explicit statement of the dimension reduction factor or the precise relation between the original and the auxiliary constrained system; this relation is load-bearing for the claim that the method reduces computational complexity.
minor comments (1)
  1. Notation for the constrained JSR and the admissible language is introduced without a dedicated preliminary section, making it difficult to track the precise dependence on the switching constraint throughout the derivations.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the careful reading and constructive comments. We address the major comments point by point below.

read point-by-point responses
  1. Referee: [Abstract] Abstract, paragraph 3: the claimed guarantee is expressed solely in terms of the p-radius and the entropy of the admissible language, yet the manuscript supplies neither an algorithm nor a reduction showing that either quantity is computable by a procedure whose complexity is strictly lower than that of the original constrained-JSR problem; without such a demonstration the bound remains non-constructive and does not tighten the practical analysis of SOS relaxations.

    Authors: The central contribution is an explicit a-priori bound on the gap between the SOS upper bound and the true constrained JSR, expressed in terms of the p-radius and the topological entropy of the admissible language. We do not claim or demonstrate that these two quantities are always computable by a strictly cheaper procedure; the bound is therefore theoretical rather than a fully constructive certificate. In many structured cases (regular languages, known entropy bounds from automata theory, or easily computable p-radius upper bounds) the quantities can be obtained independently of the full JSR. We will revise the abstract and the opening of Section 1 to state this limitation explicitly and to clarify that the result quantifies the accuracy of SOS relaxations when the auxiliary quantities are known or bounded, rather than supplying a new computational shortcut. revision: partial

  2. Referee: [Final section] The reduction for low-rank matrices (final section) is stated without an explicit statement of the dimension reduction factor or the precise relation between the original and the auxiliary constrained system; this relation is load-bearing for the claim that the method reduces computational complexity.

    Authors: We agree that the final section would benefit from greater precision. The reduction shows that the (unconstrained) JSR of a set of matrices of maximum rank r can be recovered exactly from the constrained JSR of an auxiliary system whose matrices have dimension strictly smaller than the original (the precise factor is linear in r). In the revised manuscript we will add an explicit theorem stating the dimension-reduction factor and the exact algebraic relation between the two JSR values, thereby making the complexity claim rigorous. revision: yes

Circularity Check

0 steps flagged

No circularity; derivation uses independent standard quantities

full rationale

The paper derives an upper-bound guarantee for the SOS relaxation of the constrained JSR expressed in terms of the p-radius and the topological entropy of the allowed switching language; both quantities are standard, independently defined objects in the switched-systems literature and are not shown to be computed from the same optimization as the target JSR. The final section supplies an explicit reduction from low-rank unconstrained JSR to constrained JSR of lower dimension, which is a one-way mapping rather than a definitional loop. No equations, self-citations, or fitted parameters are exhibited that would make any claimed prediction equivalent to its own inputs by construction.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Only the abstract is available; no free parameters, axioms, or invented entities can be identified.

pith-pipeline@v0.9.0 · 5677 in / 1183 out tokens · 40128 ms · 2026-05-25T12:16:13.903954+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

34 extracted references · 34 canonical work pages

  1. [1]

    The boundedness of all products of a pair of matrices is undecidable,

    V . D. Blondel and J. N. Tsitsiklis, “The boundedness of all products of a pair of matrices is undecidable,” Systems & Control Letters , vol. 41, no. 2, pp. 135–140, 2000

  2. [2]

    A note on the joint spectral radius,

    G.-C. Rota and W. Strang, “A note on the joint spectral radius,” Proceedings of the Netherlands Academy , 1960, 22:379–381

  3. [3]

    Stable adap- tive co-simulation: A switched systems approach,

    C. Gomes, B. Legat, R. M. Jungers, and H. Vangheluwe, “Stable adap- tive co-simulation: A switched systems approach,” in IUTAM Symposium on Co-Simulation and Solver Coupling , no. 1, Darmstadt, Germany, 2017, p. to appear

  4. [4]

    Jungers, The joint spectral radius: theory and applications

    R. Jungers, The joint spectral radius: theory and applications. Springer Science & Business Media, 2009, vol. 385

  5. [5]

    A Gel’fand-type spectral radius formula and stability of linear constrained switching systems,

    X. Dai, “A Gel’fand-type spectral radius formula and stability of linear constrained switching systems,” Linear Algebra and its Applications , vol. 436, no. 5, pp. 1099–1113, 2012

  6. [6]

    Quantized feedback stabilization of linear systems,

    R. W. Brockett and D. Liberzon, “Quantized feedback stabilization of linear systems,” IEEE transactions on Automatic Control, vol. 45, no. 7, pp. 1279–1289, 2000

  7. [7]

    A new method for sta- bilization of networked control systems with random delays,

    L. Zhang, Y . Shi, T. Chen, and B. Huang, “A new method for sta- bilization of networked control systems with random delays,” IEEE Transactions on automatic control, vol. 50, no. 8, pp. 1177–1181, 2005

  8. [8]

    Coordination of groups of mobile au- tonomous agents using nearest neighbor rules,

    A. Jadbabaie, J. Lin et al. , “Coordination of groups of mobile au- tonomous agents using nearest neighbor rules,” IEEE Transactions onAutomatic Control, vol. 48, no. 6, pp. 988–1001, 2003

  9. [9]

    Joint spectral radius and path-complete graph Lyapunov functions,

    A. A. Ahmadi, R. M. Jungers, P. A. Parrilo, and M. Roozbehani, “Joint spectral radius and path-complete graph Lyapunov functions,” SIAM Journal on Control and Optimization, vol. 52, no. 1, pp. 687–717, 2014

  10. [10]

    Minimally constrained stable switched systems and application to co-simulation,

    C. Gomes, R. M. Jungers, B. Legat, and H. Vangheluwe, “Minimally constrained stable switched systems and application to co-simulation,” in 57th IEEE Conference on Decision and Control . IEEE, 2018

  11. [11]

    Stability of discrete-time switching systems with constrained switching sequences,

    M. Philippe, R. Essick, G. E. Dullerud, and R. M. Jungers, “Stability of discrete-time switching systems with constrained switching sequences,” Automatica, vol. 72, pp. 242–250, 2016

  12. [12]

    Approximation of the joint spectral radius using sum of squares,

    P. A. Parrilo and A. Jadbabaie, “Approximation of the joint spectral radius using sum of squares,” Linear Algebra and its Applications , vol. 428, no. 10, pp. 2385–2402, 2008

  13. [13]

    Tabuada, Verification and control of hybrid systems: a symbolic approach

    P. Tabuada, Verification and control of hybrid systems: a symbolic approach. Springer Science & Business Media, 2009

  14. [14]

    Lind and B

    D. Lind and B. Marcus, An introduction to symbolic dynamics and coding. Cambridge university press, 1995

  15. [15]

    Entropy for group endomorphisms and homogeneous spaces,

    R. Bowen, “Entropy for group endomorphisms and homogeneous spaces,” Transactions of the American Mathematical Society , vol. 153, pp. 401–414, 1971

  16. [16]

    Topological entropy,

    R. L. Adler, A. G. Konheim, and M. H. Konheim, “Topological entropy,” Transactions of the American Mathematical Society , vol. 114, no. 2, pp. 309–319, 1965

  17. [17]

    The kullback-leibler rate pseudo-metric for comparing dynamical systems,

    S. Yu and P. G. Mehta, “The kullback-leibler rate pseudo-metric for comparing dynamical systems,” IEEE Transactions on Automatic Con- trol, vol. 55, no. 7, pp. 1585–1598, 2010

  18. [18]

    A generalized entropy criterion for nevanlinna-pick interpolation with degree constraint,

    C. I. Byrnes, T. T. Georgiou, and A. Lindquist, “A generalized entropy criterion for nevanlinna-pick interpolation with degree constraint,” IEEE Transactions on Automatic Control , vol. 46, no. 6, pp. 822–839, 2001

  19. [19]

    Hellinger versus kullback– leibler multivariable spectrum approximation,

    A. Ferrante, M. Pavon, and F. Ramponi, “Hellinger versus kullback– leibler multivariable spectrum approximation,” IEEE Transactions on Automatic Control, vol. 53, no. 4, pp. 954–967, 2008

  20. [20]

    On the geometry of maximum entropy problems,

    M. Pavon and A. Ferrante, “On the geometry of maximum entropy problems,” SIAM Review, vol. 55, no. 3, pp. 415–439, 2013. [Online]. Available: https://doi.org/10.1137/120862843

  21. [21]

    Joint spectral radius of rank one matrices and the maximum cycle mean problem

    A. A. Ahmadi and P. A. Parrilo, “Joint spectral radius of rank one matrices and the maximum cycle mean problem.” in CDC, 2012, pp. 731–733

  22. [22]

    An entropy-based bound for the computational complexity of a switched system,

    B. Legat, P. A. Parrilo, and R. M. Jungers, “An entropy-based bound for the computational complexity of a switched system,” https://www. codeocean.com/, June 2019

  23. [23]

    Julia: A fresh approach to numerical computing,

    J. Bezanson, A. Edelman, S. Karpinski, and V . B. Shah, “Julia: A fresh approach to numerical computing,” SIAM review , vol. 59, no. 1, pp. 65–98, 2017

  24. [25]

    blegat/HybridSystems.jl: v0.3.0,

    B. Legat, M. Forets, and C. Schilling, “blegat/HybridSystems.jl: v0.3.0,” May 2019. [Online]. Available: https://doi.org/10.5281/zenodo.1246104

  25. [26]

    Sum-of-squares optimization in Julia,

    B. Legat, C. Coey, R. Deits, J. Huchette, and A. Perry, “Sum-of-squares optimization in Julia,” in The First Annual JuMP-dev Workshop , 2017

  26. [27]

    Set Programming with JuMP,

    B. Legat, R. M. Jungers, P. A. Parrilo, and P. Tabuada, “Set Programming with JuMP,” in The Third Annual JuMP-dev Workshop , 2019

  27. [28]

    JuMP: A modeling language for mathematical optimization,

    I. Dunning, J. Huchette, and M. Lubin, “JuMP: A modeling language for mathematical optimization,” SIAM Review, vol. 59, no. 2, pp. 295–320, 2017

  28. [29]

    Mosek optimization suite release 8.1.0.67,

    M. ApS, “Mosek optimization suite release 8.1.0.67,” URL: http://docs. mosek.com/8.1/intro.pdf, 2017

  29. [30]

    Generating unstable trajectories for Switched Systems via Dual Sum-Of-Squares techniques,

    B. Legat, R. M. Jungers, and P. A. Parrilo, “Generating unstable trajectories for Switched Systems via Dual Sum-Of-Squares techniques,” in Proceedings of the 19th International Conference on Hybrid Systems: Computation and Control , ser. HSCC ’16. ACM, 2016, pp. 51–60. [Online]. Available: http://doi.acm.org/10.1145/2883817.2883821

  30. [31]

    Certifying unstability of Switched Systems using Sum of Squares Programming,

    B. Legat, P. A. Parrilo, and R. M. Jungers, “Certifying unstability of Switched Systems using Sum of Squares Programming,” ArXiv e-prints, Oct. 2017

  31. [32]

    The p-norm joint spectral radius and its applications in wavelet analysis,

    D.-X. Zhou, “The p-norm joint spectral radius and its applications in wavelet analysis,” AMS IP Studies in Advanced Mathematics , vol. 25, pp. 305–326, 2002

  32. [33]

    Computationally efficient approxima- tions of the joint spectral radius,

    V . D. Blondel and Y . Nesterov, “Computationally efficient approxima- tions of the joint spectral radius,” SIAM Journal on Matrix Analysis and Applications, vol. 27, no. 1, pp. 256–272, 2005. 8

  33. [34]

    Efficient method for computing lower bounds on the p-radius of switched linear systems,

    M. Ogura, V . M. Preciado, and R. M. Jungers, “Efficient method for computing lower bounds on the p-radius of switched linear systems,” Systems & Control Letters , vol. 94, pp. 159–164, 2016

  34. [35]

    Boyd and L

    S. Boyd and L. Vandenberghe, Convex optimization . Cambridge university press, 2004