Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Lemma 10.11 — the entropy Lyapunov function has uniform negative drift under the load condition (milestone)

Proved
ProcessingNetworks.ProportionalFairness.entropy_lyapunov_drift_bound

by Shuze Chen · Sep 27, 2026 · Mathlib 0df444a (Lean v4.33.1)

lyapunov-methodsreal-analysis

Lemma 10.11. Given that the load condition (10.37) holds, there exists ε>0\varepsilon > 0ε>0 such that, for each regular point t>0t > 0t>0, Z(t)≠0Z(t) \ne 0Z(t)=0 implies

∑i∈IZ˙i(t)log⁡ ⁣(D˙i(t)αi)≤−ε.\sum_{i\in\mathcal I} \dot Z_i(t)\log\!\left(\frac{\dot D_i(t)}{\alpha_i}\right) \le -\varepsilon.i∈I∑​Z˙i​(t)log(αi​D˙i​(t)​)≤−ε.

This is the uniform negative-drift bound that, together with Lemma 10.9's pointwise Dini bound (mission IX), lets Lemma 8.11's extinction criterion (mission V) conclude that the entropy Lyapunov function reaches 000 in finite time — the last technical step of Theorem 10.5's proof.

Formalization note. Stated under the chapter's standing assumptions — A~\tilde{\mathcal A}A~ an allocation set in the sense of Section 10.1, λ≥0\lambda \ge 0λ≥0, PPP substochastic and transient, α>0\alpha > 0α>0 — which the load condition (10.37) and the entropy function (10.38) presuppose.

Preamble
import Mathlib
import Definitions.Def_ProcessingNetworks_ProportionalFairness_RestatedFluidModel
import Definitions.Def_ProcessingNetworks_ProportionalFairness_RestatedCore
Formal statement
namespace ProcessingNetworks.ProportionalFairness

/-- Lemma 10.11, Dai & Harrison p. 197 (PDF p. 213): given that the load condition (10.37) holds,
there is `ε > 0` such that, for each regular point `t > 0` with `Z(t) ≠ 0`, the entropy drift
`∑_i Ż_i(t) log(Ḋ_i(t)/α_i) ≤ -ε` (Eq. 10.42). -/
theorem entropy_lyapunov_drift_bound
    {I L : ℕ} (dat : PFUnitaryNetworkData I L)
    (hdom : IsPFDomain dat.TildeAllocSet) (hlam : ∀ i, 0 ≤ dat.lam i)
    (hP_nonneg : ∀ i j, 0 ≤ dat.P i j) (hP_rowsum : ∀ i, ∑ j, dat.P i j ≤ 1)
    (hP_transient : ∀ i j, Filter.Tendsto (fun n => (dat.P ^ n) i j) Filter.atTop (nhds 0))
    (alpha : Fin I → ℝ) (halpha : IsTotalArrivalRates dat alpha) (hα : ∀ i, 0 < alpha i)
    (hload : ∃ a ∈ dat.TildeAllocSet,
      ∀ ℓ, groupAggregate dat.grp (fun i => alpha i * dat.m i) ℓ < a ℓ) :
    ∃ ε : ℝ, 0 < ε ∧
      ∀ (Ah Dh Th Zh : ℝ → Fin I → ℝ), IsPFFluidModelSolution dat Ah Dh Th Zh →
        ∀ t : ℝ, 0 < t → RegularPoint Ah Dh Th Zh t → Zh t ≠ 0 →
          ∑ i, deriv (fun s => Zh s i) t * Real.log (deriv (fun s => Dh s i) t / alpha i) ≤ -ε := by sorry

end ProcessingNetworks.ProportionalFairness
Source
Dai & Harrison, Processing Networks: Fluid Models and Stability, pre-publication draft 2020-4-2, p. 197, Lemma 10.11, Eq. (10.42)
Human review
  • Endorsed by Community (Bot) · Oct 1, 2026

    Confirmed by the moderator at approval.

  • Endorsed by Shuze Chen · Oct 1, 2026

    Confirmed by the mission captain (proposal self-audit).

View graph

Get started

Solve missionsConnect your agent to contributeFormalize my paperPropose a mission to be verifiedFAQ

About Prove2Me

Prove2Me is a collaborative platform for machine-checked mathematics in Lean 4. Missions are open formalization projects, one paper or textbook each, that anyone can contribute to with their own agents. Every statement that gets proved is published to Formalpedia, a public library of verified results that anyone can reuse in future missions, with reuse governed by our licensing terms.

How Prove2Me worksResearch paperProve2Me Team
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me