Theorem 2.2.1 — the queueing proof of the pointwise ergodic theorem
ProvedPalmQueueing.Loynes.queueing_ergodic_theoremTheorem 2.2.1. Whenever is ergodic and both and are non-negative, not identically null, and integrable,
The direction of the argument is the point, and the section's title says so: "Queueing Proof of the Ergodic Theorem". The queueing construction of §2.2 — Loynes' monotone scheme, Hopf's lemma (2.2.18) and the stability criterion — yields the pointwise ergodic theorem rather than being derived from it. The ratio form is what that route naturally produces, since what the queue supplies is the comparison with whenever ; a scaling argument over all with then gives the bound , and exchanging the roles of and gives the matching .
Mathlib's Ergodic is reused rather than a private notion of ergodicity. Mathlib has the mean
(von Neumann) ergodic theorem but no pointwise one, so nothing here duplicates an existing
result.
import Mathlib /-! # Theorem 2.2.1: the queueing proof of the pointwise ergodic theorem (§2.2.5, p.90) -/
namespace PalmQueueing.Loynes
open MeasureTheory Filter Topology
variable {Ω : Type*} [MeasurableSpace Ω]
/-- **Theorem 2.2.1** (§2.2.5, p.90). Whenever `(P⁰, θ)` is ergodic and both `σ` and `τ` are
non-negative, not identically null, and integrable,
`lim_{n→∞} ( Σ_{i=0}^{n} σ ∘ θ^{-i} ) / ( Σ_{i=0}^{n} τ ∘ θ^{-i} ) = E⁰[σ] / E⁰[τ]`, `P⁰`-a.s.
The section's title is "Queueing Proof of the Ergodic Theorem", and the point is the direction of
the argument: the queueing construction of §2.2 — Loynes' monotone scheme, Hopf's lemma (2.2.18)
and the stability criterion — *yields* the pointwise ergodic theorem, rather than being derived
from it. The ratio form is the natural output of that route, since what the queue supplies is the
comparison `Σ σ ∘ θ^{-i} ≤ Σ τ ∘ θ^{-i} + M_n` with `M_n ↑ M_∞ < ∞`.
Mathlib's `Ergodic` is reused rather than a private notion; Mathlib has the *mean* ergodic theorem
but no pointwise one, so nothing here duplicates it. -/
theorem queueing_ergodic_theorem (P0 : Measure Ω) [IsProbabilityMeasure P0]
(shift : Ω ≃ᵐ Ω) (herg : Ergodic shift P0)
(sig tau : Ω → ℝ) (hsigmeas : Measurable sig) (htaumeas : Measurable tau)
(hsig0 : ∀ ω, 0 ≤ sig ω) (htau0 : ∀ ω, 0 ≤ tau ω)
(hsigNull : ¬ (∀ᵐ ω ∂P0, sig ω = 0)) (htauNull : ¬ (∀ᵐ ω ∂P0, tau ω = 0))
(hsigInt : Integrable sig P0) (htauInt : Integrable tau P0) :
∀ᵐ ω ∂P0, Tendsto
(fun n : ℕ =>
(∑ i ∈ Finset.range (n + 1), sig ((shift.symm^[i]) ω)) /
(∑ i ∈ Finset.range (n + 1), tau ((shift.symm^[i]) ω)))
atTop (𝓝 ((∫ ω, sig ω ∂P0) / ∫ ω, tau ω ∂P0)) := by sorry
end PalmQueueing.Loynes
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.