Lyapunov Exponent of Rank One Matrices: Ergodic Formula and Inapproximability of the Optimal Distribution

preprint2019arXivOpen access

Abstract

The Lyapunov exponent corresponding to a set of square matrices A={A1,…,An}\mathcal{A} = \{A_1, \dots, A_n \} and a probability distribution pp over {1,…,n}\{1, \dots, n\} is λ(A,p):=lim⁡k→∞1k Elog⁡∥Aσk⋯Aσ2Aσ1∥λ(\mathcal{A},p) := \lim_{k \to \infty} \frac{1}{k} \,\mathbb{E} \log \|A_{σ_k} \cdots A_{σ_2}A_{σ_1}\|, where σiσ_i are i.i.d. according to pp. This quantity is of fundamental importance to control theory since it determines the asymptotic convergence rate eλ(A,p)e^{λ(\mathcal{A},p)} of the stochastic linear dynamical system xk+1=Aσkxkx_{k+1} = A_{σ_k} x_k. This paper investigates the following "design problem": given A\mathcal{A}, compute the distribution pp minimizing λ(A,p)λ(\mathcal{A},p). Our main result is that it is NP-hard to decide whether there exists a distribution pp for which λ(A,p)<0λ(\mathcal{A},p)< 0, i.e. it is NP-hard to decide whether this dynamical system can be stabilized. This hardness result holds even in the "simple"' case where A\mathcal{A} contains only rank-one matrices. Somewhat surprisingly, this is in stark contrast to the Joint Spectral Radius -- the deterministic kindred of the Lyapunov exponent -- for which the analogous optimization problem for rank-one matrices is known to be exactly computable in polynomial time. To prove this hardness result, we first observe via Birkhoff's Ergodic Theorem that the Lyapunov exponent of rank-one matrices admits a simple formula and in fact is a quadratic form in pp. Hardness of the design problem is shown through a reduction from the Independent Set problem. Along the way, simple examples are given illustrating that p↦λ(A,p)p \mapsto λ(\mathcal{A},p) is neither convex nor concave in general. We conclude with extensions to continuous distributions, exchangeable processes, Markov processes, and stationary ergodic processes.

Explore this paper’s authors, topics and related work.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

Reviews 0

Write a reviewWrite

No reviews yet.

Discussion 0

Add a commentWrite

No comments yet.