Proposition 2.1 (p. 8) — Σ_{i≤A(t)} (G(t − τᵢ − wᵢ) − G(t − τᵢ)) = ∫₀ᵗ (Q(t − s) − N)⁺ dF(s) − Σ (G(t − w̃ᵢ) − G(t))
ProvedReedGGN.SystemEq.proposition_2_1Let be a service-time distribution on with tail , and fix a non-idling sample path of the queue: servers, initial customers, arrival times with counting process , waiting times of the arrivals and of the initial customers , and number in system given by (2.2). Then for each ,
The left side is the correction that the waiting times make to the infinite-server terms in (2.7). The proposition expresses it through the number of waiting customers alone, which is what turns (2.7) into the closed system equation (2.8).
Formalization Note "Non-idling" is the identity for all , with which the paper's proof opens. The integral is over the closed interval , an atom of at included. is a probability measure on with . The paper's assumption that has mean is not used and is omitted, so the statement holds for every service-time distribution on .
import Mathlib import Definitions.Def_ReedGGN_SystemEq_NonIdling
namespace ReedGGN.SystemEq
open MeasureTheory
/-- Proposition 2.1 (p. 8): for a non-idling sample path and each `t ≥ 0`,
`Σ_{i=1}^{A(t)} (G(t − τ_i − w_i) − G(t − τ_i))
= ∫_0^t (Q(t − s) − N)⁺ dF(s) − Σ_{i=1}^{(Q₀−N)⁺} (G(t − w̃_i) − G(t))`,
the integral over the closed interval `[0, t]`. -/
theorem proposition_2_1 (μ : Measure ℝ) [IsProbabilityMeasure μ] (hμ : μ (Set.Iio 0) = 0)
(P : SamplePath) (hP : NonIdling P) (t : ℝ) (ht : 0 ≤ t) :
∑ i ∈ Finset.Icc 1 (P.A t), (G μ (t - P.τ i - P.w i) - G μ (t - P.τ i)) =
(∫ s in Set.Icc 0 t, max ((Q P (t - s) : ℝ) - P.N) 0 ∂μ) -
∑ i ∈ Finset.Icc 1 (P.Q₀ - P.N), (G μ (t - P.wt i) - G μ t) := by sorry
end ReedGGN.SystemEq
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.