Absolutely continuous invariant measures for expanding piecewise linear maps (Q5927641): Difference between revisions

Suppose that \(U\) is a (not necessarily bounded or connected) polyhedron in \({\mathbb{R}}^{d}\), and let \({\mathcal P}\) be a finite family of polyhedra whose interiors are pairwise disjoint. A map \({\mathcal T}:U\to U\) is called piecewise linear, if for each \(P\in{\mathcal P}\) the restriction of \({\mathcal T}\) to the interior of \(P\) is an affine map \(A_{P}x+v_{P}\). If for every \(P\in{\mathcal P}\) all eigenvalues of \(A_{P}\) have modulus strictly larger than \(1\), then \({\mathcal T}\) is called expanding. The author proves in this paper that every expanding piecewise linear map \({\mathcal T}\) on a bounded polyhedron has an invariant probability measure, which is absolutely continuous with respect to the Lebesgue measure. Moreover, there exist only finitely many ergodic absolutely continuous invariant probability measures \(\mu_{1},\mu_{2},\dots ,\mu_{p}\). Furthermore there exist open sets \(U_{1},U_{2},\dots ,U_{p}\) with \(\bigcup_{k=1}^{p}U_{k}\) has full Lebesgue measure in \(U\), such that \(\frac{1}{n}\sum_{j=1}^{n-1} \delta_{{\mathcal T}^{j}(x)}\) converges weakly to \(\mu_{k}\) for every \(k\) and each \(x\in U_{k}\). As in \textit{F. Hofbauer} and \textit{G. Keller} [Math. Z. 180, 119-140 (1982; Zbl 0485.28016)] one can derive strong ergodic properties of the \(\mu_{k}\) (weak Bernoullicity, exponential decay of correlation, central limit theorems). It is also proved that the associated Perron-Frobenius operator on the space of functions of bounded variation is quasi compact and its essential spectral radius is bounded by a certain expansion constant of \({\mathcal T}\). Similar results are known for other types of dynamical systems. These are generalizations of the classical result obtained by \textit{A. Lasota} and \textit{J. Yorke} [Trans. Am. Math. Soc. 186, 481-488 (1973; Zbl 0298.28015)] in the one-dimensional case for a class of maps much larger than the family of piecewise linear maps. In the proofs a quantity called ``weighted multiplicity'' is used. Using delicate estimations the author shows a result on the spectrum of the associated Perron-Frobenius operator on the space of functions of bounded variation. From this result the existence of absolutely continuous invariant probability measures and their ergodic properties can be derived in the classical way. Besides of presenting interesting new results this paper is also written very well. It provides an important step towards the understanding of absolutely continuous invariant measures in the more-dimensional case.