Theorem 4.3 — the FBFS fluid model of every reentrant line is stable
ProvedDaiWeissFluid.FBFS.theorem4_3Consider any reentrant line with stations and classes, mean service times , and nominal workloads satisfying
Then the fluid model corresponding to the First-Buffer-First-Served discipline, i.e. (1.8)–(1.12) together with the preemptive-resume priority condition (4.4) for , is stable: there is a time such that every solution with satisfies for all and all classes .
By the paper's Theorem 1.1 (Dai 1995), stability of this fluid model implies positive Harris recurrence of the corresponding multiclass queueing network under FBFS whenever (1.7) holds, the stochastic analogue of Kumar's result for deterministic systems.
Formalization Note The statement quantifies over every line: all , , routes and service times . The positivity is explicit (the paper takes it for granted, since ). Condition (1.13) is not added, since (4.4) supersedes it (p. 123). Stability is Definition 1.3, FluidStable. Classes and stations are 0-based.
import Mathlib import Definitions.Def_DaiWeissFluid_FBFS_FluidModel
namespace DaiWeissFluid.FBFS
/-- Theorem 4.3, p. 124: in every reentrant line satisfying (1.7), the fluid model of the
First-Buffer-First-Served discipline, (1.8)–(1.12) with (4.4) for `π(k) = k`, is stable
(Definition 1.3). -/
theorem theorem4_3 {I K : ℕ} (L : ReentrantLine I K) (hm : ∀ k, 0 < L.m k)
(hρ : ∀ i, L.ρ i < 1) :
DaiWeissFluid.ThreeBuffer.FluidStable (L.IsPrioritySolution (fbfs K)) := by sorry
end DaiWeissFluid.FBFS
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.