Eq. (7.12) —
ProvedQueueingFundamentals.Bounds.departure_variancedeparture-processgg1p2o-batch-b23bp2o-gran-per-chapterp2o-plan-bookp2o-v1queueing
Consider a G/G/1 queue whose interarrival time and service time are nonnegative, have finite second moments, and satisfy , , and , with and . Let be a stationary law of the line delay under Lindley's recursion (7.1). The interdeparture time is
where , , and are independent.
Then is finite and the interdeparture time in steady state satisfies
This is the variance of the departure process of a stationary G/G/1 queue. The approximations of §7.3 for queueing networks use it.
Formalization Note is a function on the product space of with law . The variance is Mathlib's variance.
Preamble
import Mathlib import Definitions.Def_QueueingFundamentals_Bounds_GG1
Formal statement
namespace QueueingFundamentals.Bounds
open MeasureTheory ProbabilityTheory Filter Topology
/-- Eq. (7.12), p.332. For a stationary G/G/1 queue with `ρ < 1`, the interdeparture time
`D = S^{(n+1)} + X^{(n)}` (with `S^{(n+1)} ~ B` independent of `(W_q^{(n)}, S^{(n)}, T^{(n)})`)
satisfies `Var[D] = 2σ_B² + σ_A² − 2W_q(1/λ − 1/μ)`. -/
theorem departure_variance
{A B ν : Measure ℝ} {lam mu : ℝ} (hlam : 0 < lam) (hmu : 0 < mu)
(hin : IsGG1Input A B lam mu) (hrho : lam / mu < 1) (hν : IsStationaryWaitLaw A B ν) :
Integrable (fun w : ℝ => w) ν ∧
variance (fun q : ℝ × (ℝ × ℝ × ℝ) => q.1 + idleX q.2.1 q.2.2.1 q.2.2.2)
(B.prod (stepLaw ν B A)) =
2 * serviceVar B + interarrivalVar A - 2 * meanWait ν * (1 / lam - 1 / mu) := by sorry
end QueueingFundamentals.Bounds
Source
Gross, Shortle, Thompson & Harris, Fundamentals of Queueing Theory, 4th ed., Wiley 2008, DOI 10.1002/9781118625651, pp.331–332, D^(n) ≡ S^(n+1) + X^(n) and Eq. (7.12)
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.