Eq. (3.55): the partial generating functions of the M/M/1 retrial queue in closed form
ProvedQueueingFundamentals.AdvMarkov.retrial_pgf_closed_formgenerating-functionp2o-batch-b23bp2o-gran-per-chapterp2o-plan-bookp2o-v1queueingretrial-queue
Consider the retrial queue with arrival rate , service rate and retrial rate , and let . Let be a steady-state solution of the rate-balance equations (3.47)–(3.49), with partial generating functions and . Then for every real with ,
At this gives , the fraction of time the server is busy. Expanding these functions in powers of yields the steady-state probabilities (3.57).
Formalization Note The book does not state the range of ; the statement takes real , where the series converge absolutely and . The power is a real power of a positive number.
Preamble
import Mathlib import Definitions.Def_QueueingFundamentals_AdvMarkov_Retrial
Formal statement
namespace QueueingFundamentals.AdvMarkov
/-- Eq. (3.55), p.161: for a steady-state solution of the `M/M/1` retrial queue with
`ρ = λ/μ < 1`, the partial generating functions are, for `z ∈ [−1, 1]`,
`P_0(z) = (1 − ρz)((1 − ρ)/(1 − ρz))^{(λ/γ)+1}` and `P_1(z) = ρ((1 − ρ)/(1 − ρz))^{(λ/γ)+1}`. -/
theorem retrial_pgf_closed_form (lam mu gam : ℝ) (hlam : 0 < lam) (hmu : 0 < mu) (hgam : 0 < gam)
(ρ : ℝ) (hρ : ρ = lam / mu) (hρ1 : ρ < 1)
(p0 p1 : ℕ → ℝ) (hp : IsRetrialSteadyState lam mu gam p0 p1) :
∀ z ∈ Set.Icc (-1 : ℝ) 1,
pgf p0 z = (1 - ρ * z) * ((1 - ρ) / (1 - ρ * z)) ^ (lam / gam + 1) ∧
pgf p1 z = ρ * ((1 - ρ) / (1 - ρ * z)) ^ (lam / gam + 1) := by sorry
end QueueingFundamentals.AdvMarkov
Source
Gross, Shortle, Thompson & Harris, Fundamentals of Queueing Theory, 4th ed., Wiley 2008, DOI 10.1002/9781118625651, p.161, Eq. (3.55)
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.