Exercise 1.9(ii): exact upward passage probability for the overloaded M/G/1 surplus
ProvedKyprianou.Ch01.queueUpperPassageProbabilityexercise-1-9first-passagekyprianou-ch01queueing
Consider the M/G/1 / Cramér–Lundberg process with unit drift,
built from a compound Poisson driver with arrival rate and i.i.d. service times (claims) . Assume:
- is concentrated on , i.e. ;
- has finite mean and the queue is overloaded, ;
- is a root of the Laplace exponent
Then for every ,
Equivalently, the overall supremum is exponentially distributed with parameter . This is the second assertion of Kyprianou's Exercise 1.9(ii); together with the exponential martingale (the first assertion) it describes the upward fluctuations of the overloaded M/G/1 surplus, which has no positive jumps and drifts to .
Formalization Note. is riskSurplus arrival claims 1 0 and is queueLaplaceExponent rate F. The probability is the real value of the (outer) measure of the strict upward passage event.
Preamble
import Mathlib import Definitions.Def_Kyprianou_Ch01_RiskQueueCore import Definitions.Def_Kyprianou_Ch01_McKeanBrownianStopping import Definitions.Def_LeanLevy_Processes_CompoundPoisson open MeasureTheory ProbabilityTheory Kyprianou.Ch01 Set Filter open scoped NNReal ENNReal BigOperators set_option autoImplicit false
Formal statement
theorem Kyprianou.Ch01.queueUpperPassageProbability
{Ω : Type*} [MeasurableSpace Ω]
(P : Measure Ω) [IsProbabilityMeasure P]
(arrival claims : ℕ → Ω → ℝ)
(rate : ℝ≥0) (hrate : 0 < rate)
(F : Measure ℝ) [IsProbabilityMeasure F]
(hDriver : IsCompoundPoissonDriver arrival claims rate F P)
(hPos : F (Set.Iic (0 : ℝ)) = 0)
(hIntegrable : Integrable (fun z : ℝ => z) F)
(mu : ℝ) (hMean : (∫ z, z ∂F) = mu)
(hOverload : 1 < (rate : ℝ) * mu)
(theta : ℝ) (hTheta : 0 < theta)
(hRoot : queueLaplaceExponent rate F theta = 0) :
∀ x : ℝ, 0 < x →
(P {ω | ∃ t : ℝ≥0, 0 < t ∧ x < riskSurplus arrival claims 1 0 t ω}).toReal =
Real.exp (-theta * x) := by
sorrySource
Kyprianou (2006), Introductory Lectures on Fluctuations of Lévy Processes with Applications, Chapter 1, Exercise 1.9(ii), printed p. 31 (second assertion: the passage probability).