A Ruelle-Perron-Frobenius theorem for expanding circle maps with an indifferent fixed point (Q2674352)
From MaRDI portal
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | A Ruelle-Perron-Frobenius theorem for expanding circle maps with an indifferent fixed point |
scientific article |
Statements
A Ruelle-Perron-Frobenius theorem for expanding circle maps with an indifferent fixed point (English)
0 references
12 September 2022
0 references
Let \(\mathbb{T} = \mathbb{R}/\mathbb{Z} = [0, 1)\) be the circle equipped with the standard metric. The authors consider a family \(\mathcal{F}\) of nonuniformly expanding continuous maps with an indifferent fixed point. That is, maps \(T : \mathbb{T} \to \mathbb{T}\) given by the expression \(T(x) := x(1 + V(x)) \bmod 1\), with \(V : [0, \infty) \to [0, \infty)\) an increasing function satisfying that \(V(1) \in \mathbb{N}\) and \(\lim_{x \to 0}\frac{V(tx)}{V(x)} = t^\sigma\) for each \(t > 0\), where \(\sigma \in [0, \infty)\). Given \(T \in \mathcal{F}\), the authors present a notion of \(T\)-compatibility between any pair of concave modulus of continuity. Moreover, fixing \(\omega\) and \(\Omega\) two \(T\)-compatible modulus of continuity, they prove that the Ruelle operator \(\mathcal{L}_f\) associated to a map \(f \in \mathcal{C}_\omega(\mathbb{T})\) satisfies a Ruelle-Perron-Frobenius theorem in the following sense: there exist a positive constant \(\chi\), a Borel probability measure \(\nu\) and a strictly positive function \(h \in \mathcal{C}_\Omega(\mathbb{T})\) such that \(\mathcal{L}_f^* \nu = \chi \nu\), \(\mathcal{L}_f h = \chi h\), \(\int h d\nu = 1\) and the sequence \((\chi^{-n}\mathcal{L}_f^n \phi)_{n \in \mathbb{N}}\) converges uniformly to \(\int \phi d\nu\) on \(\mathbb{T}\). They also prove that the \(T\)-invariant probability measure \(d\mu := h d\nu\) satisfies a variational principle of the pressure, i.e., \[ h_\mu(T) + \int f d\mu = \log \chi = P(T, f), \] where \[ P(T, f) := \sup\Bigl\{ h_m(T) + \int f dm : m \text{ is }T\text{-invariant} \Bigr\}. \] Besides that, they prove that \(d\mu := h d\nu\) results in a Gibbs measure, in the sense that given \(r > 0\) small enough, there is \(K_r > 0\) such that for any \(x \in \mathbb{T}\) and any \(n \geq 1\), there holds \[ K_r^{-1} \leq \frac{\mu(\{y : d(T^j(x), T^j(y)) < r, 0 \leq j \leq n\})}{e^{S_n f(x) - nP(T, f)}} \leq K_r. \] Finally, they present some interesting corollaries in the setting of Manneville-Pomeau maps.
0 references
nonuniformly expanding dynamics
0 references
intermittent maps
0 references
thermodynamic formalism
0 references
equilibrium states
0 references
modulus of continuity
0 references
0 references
0 references
0 references
0 references