bounds for the coupling time in queueing networks perfect ... · outline 1 queueing networks with...

79
Bounds for the Coupling Time in Queueing Networks Perfect Simulation J.G. Dopper 2 , B. Gaujal and J.-M. Vincent 1 1 Laboratory ID-IMAG MESCAL Project Universities of Grenoble, France {Bruno.Gaujal,Jean-Marc.Vincent}@imag.fr 2 Mathematical Institute, Leiden University, Nederland [email protected] J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble) Bounds for the Coupling Time in Queueing Networks Perfect Simulation MAM 2006, june12 1 / 27

Upload: others

Post on 24-Oct-2020

5 views

Category:

Documents


0 download

TRANSCRIPT

  • Bounds for the Coupling Time in QueueingNetworks Perfect Simulation

    J.G. Dopper2, B. Gaujal and J.-M. Vincent1

    1Laboratory ID-IMAGMESCAL Project

    Universities of Grenoble, France{Bruno.Gaujal,Jean-Marc.Vincent}@imag.fr

    2Mathematical Institute,Leiden University, [email protected]

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 1 / 27

  • Outline

    1 Queueing Networks with finite capacity

    2 Event modelling and monotonicity

    3 Perfect simulation and coupling time

    4 Acyclic networks

    5 Synthesis and future works

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 2 / 27

  • Outline

    1 Queueing Networks with finite capacity

    2 Event modelling and monotonicity

    3 Perfect simulation and coupling time

    4 Acyclic networks

    5 Synthesis and future works

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 3 / 27

  • Queueing networks with finite capacity

    Network modelFinite set of resources :

    servers

    waiting room

    Routing strategies :

    state dependent

    overflow strategy

    blocking strategy...

    Average performance :

    load of the system

    response time

    loss rate ...

    Markov model

    Assumptions :- Poisson arrival,- exponential distribution for service times,- probabilistic routing with overflow

    ⇒ continuous time Markov chain

    ProblemComputation of the stationary distribution⇒ state space explosion

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 4 / 27

  • Queueing networks with finite capacity

    Network modelFinite set of resources :

    servers

    waiting room

    Routing strategies :

    state dependent

    overflow strategy

    blocking strategy...

    Average performance :

    load of the system

    response time

    loss rate ...

    Markov model

    5

    C

    C

    C

    C0

    1

    2

    λ

    λ

    λ

    λλ

    01

    2

    3

    4

    Assumptions :- Poisson arrival,- exponential distribution for service times,- probabilistic routing with overflow

    ⇒ continuous time Markov chain

    ProblemComputation of the stationary distribution⇒ state space explosion

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 4 / 27

  • Queueing networks with finite capacity

    Network modelFinite set of resources :

    servers

    waiting room

    Routing strategies :

    state dependent

    overflow strategy

    blocking strategy...

    Average performance :

    load of the system

    response time

    loss rate ...

    Markov model

    5

    C

    C

    C

    C0

    1

    2

    λ

    λ

    λ

    λλ

    01

    2

    3

    4

    Assumptions :- Poisson arrival,- exponential distribution for service times,- probabilistic routing with overflow

    ⇒ continuous time Markov chain

    ProblemComputation of the stationary distribution⇒ state space explosion

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 4 / 27

  • Related works

    Non reversible systems (reverse event)Product form solution ??Widely studied domain

    - Analytical solution [Perros 94]- specific cases- numerical computation of normalization constant

    - Numerical computation [Stewart 94]

    - Approximation techniques [Onvural 90, Perros 94,...]

    - Simulation [Banks & al. 01,...]simulation of Markov modelssimulation of event graphsdiscrete event simulationperfect simulation [Mattson 04]

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 5 / 27

  • Outline

    1 Queueing Networks with finite capacity

    2 Event modelling and monotonicity

    3 Perfect simulation and coupling time

    4 Acyclic networks

    5 Synthesis and future works

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 6 / 27

  • Event modelling

    Queueing model :

    5

    C

    C

    C

    C0

    1

    2

    λ

    λ

    λ

    λλ

    01

    2

    3

    4

    Event description :rate origin destination enabling condition routing policy

    e0 λ0 Q−1 Q0 none rejection if Q0 is fulle1 λ1 Q0 Q1 s0 > 0 rejection if Q1 is fulle2 λ2 Q0 Q2 s0 > 0 rejection if Q2 is fulle3 λ3 Q1 Q3 s1 > 0 rejection if Q3 is fulle4 λ4 Q2 Q3 s2 > 0 rejection if Q3 is fulle5 λ5 Q3 Q−1 s3 > 0 none

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 7 / 27

  • Event modelling

    Multidimensional state spaceX = X0 × · · · × XK−1]

    with Xi = {0, · · · , Ci}.Event e :; transition function Φ(., e);; Poisson process λe

    Poisson driven system

    Uniformization ⇒ GSMP representation

    Λ =∑

    e

    λe and P(event e) =λeΛ

    ; Trajectory : {en}n∈Z i.i.d.

    ⇒ Homogeneous Discrete Time Markov Chain [Bremaud 99]

    Xn+1 = Φ(Xn, en+1).J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 8 / 27

  • Event modelling

    Multidimensional state spaceX = X0 × · · · × XK−1]

    with Xi = {0, · · · , Ci}.Event e :; transition function Φ(., e);; Poisson process λe

    Poisson driven system

    Time

    States

    Events

    e1

    e2

    e3

    e4

    Uniformization ⇒ GSMP representation

    Λ =∑

    e

    λe and P(event e) =λeΛ

    ; Trajectory : {en}n∈Z i.i.d.

    ⇒ Homogeneous Discrete Time Markov Chain [Bremaud 99]

    Xn+1 = Φ(Xn, en+1).J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 8 / 27

  • Event modelling

    Multidimensional state spaceX = X0 × · · · × XK−1]

    with Xi = {0, · · · , Ci}.Event e :; transition function Φ(., e);; Poisson process λe

    Poisson driven system

    Time

    States

    Events

    e1

    e2

    e3

    e4

    Uniformization ⇒ GSMP representation

    Λ =∑

    e

    λe and P(event e) =λeΛ

    ; Trajectory : {en}n∈Z i.i.d.

    ⇒ Homogeneous Discrete Time Markov Chain [Bremaud 99]

    Xn+1 = Φ(Xn, en+1).J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 8 / 27

  • Monotonicity of routing strategy

    (X ,≺) partially ordered set (componentwise)

    x = [x0, x1, · · · , xK−1] ≺ y = [y0, y1, · · · , yK−1] iff ∀i , xi 6 yi .

    An event e is said to be monotone if

    x ≺ y ⇒ Φ(x , e) ≺ Φ(y , e).

    Examples [Glasserman and Yao]All of these routing events are monotone:- external arrival with overflow and rejection- routing with overflow and rejection or blocking- routing to the shortest available queue- routing to the shortest mean available response time- general index policies [Palmer-Mitrani]- rerouting inside queues...

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 9 / 27

  • Monotonicity of routing strategy

    (X ,≺) partially ordered set (componentwise)

    x = [x0, x1, · · · , xK−1] ≺ y = [y0, y1, · · · , yK−1] iff ∀i , xi 6 yi .

    An event e is said to be monotone if

    x ≺ y ⇒ Φ(x , e) ≺ Φ(y , e).

    Examples [Glasserman and Yao]All of these routing events are monotone:- external arrival with overflow and rejection- routing with overflow and rejection or blocking- routing to the shortest available queue- routing to the shortest mean available response time- general index policies [Palmer-Mitrani]- rerouting inside queues...

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 9 / 27

  • Outline

    1 Queueing Networks with finite capacity

    2 Event modelling and monotonicity

    3 Perfect simulation and coupling time

    4 Acyclic networks

    5 Synthesis and future works

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 10 / 27

  • Classical forward simulation

    ForwardRepresentation : transition fonction

    Xn+1 = Φ(Xn, en+1).

    x ← x0{choice of the initial state at time =0}n = 0;repeat

    n ← n + 1;e ← Random event();x ← Φ(x, e);{computation of the next state Xn+1}

    until some empirical criteriareturn x

    Convergence : biased sampleSampling : Warm-up period

    Trajectory

    ComplexityRelated to the stabilization periodEstimation : replication or ergodic estimation

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 11 / 27

  • Classical forward simulation

    ForwardRepresentation : transition fonction

    Xn+1 = Φ(Xn, en+1).

    x ← x0{choice of the initial state at time =0}n = 0;repeat

    n ← n + 1;e ← Random event();x ← Φ(x, e);{computation of the next state Xn+1}

    until some empirical criteriareturn x

    Convergence : biased sampleSampling : Warm-up period

    Trajectory

    ComplexityRelated to the stabilization periodEstimation : replication or ergodic estimation

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 11 / 27

  • Classical forward simulation

    ForwardRepresentation : transition fonction

    Xn+1 = Φ(Xn, en+1).

    x ← x0{choice of the initial state at time =0}n = 0;repeat

    n ← n + 1;e ← Random event();x ← Φ(x, e);{computation of the next state Xn+1}

    until some empirical criteriareturn x

    Convergence : biased sampleSampling : Warm-up period

    Trajectory

    Initial state

    0001

    0010

    0011

    0100

    0101

    0110

    1000

    1001

    1010

    1100

    0

    Time

    1 2 3 4 5 5 6 7 8

    States

    0000

    ComplexityRelated to the stabilization periodEstimation : replication or ergodic estimation

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 11 / 27

  • Classical forward simulation

    ForwardRepresentation : transition fonction

    Xn+1 = Φ(Xn, en+1).

    x ← x0{choice of the initial state at time =0}n = 0;repeat

    n ← n + 1;e ← Random event();x ← Φ(x, e);{computation of the next state Xn+1}

    until some empirical criteriareturn x

    Convergence : biased sampleSampling : Warm-up period

    Trajectory

    Initial state

    0001

    0010

    0011

    0100

    0101

    0110

    1000

    1001

    1010

    1100

    0

    Time

    1 2 3 4 5 5 6 7 8

    States

    0000

    ComplexityRelated to the stabilization periodEstimation : replication or ergodic estimation

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 11 / 27

  • Classical forward simulation

    ForwardRepresentation : transition fonction

    Xn+1 = Φ(Xn, en+1).

    x ← x0{choice of the initial state at time =0}n = 0;repeat

    n ← n + 1;e ← Random event();x ← Φ(x, e);{computation of the next state Xn+1}

    until some empirical criteriareturn x

    Convergence : biased sampleSampling : Warm-up period

    Trajectory

    Initial state

    0001

    0010

    0011

    0100

    0101

    0110

    1000

    1001

    1010

    1100

    0

    Time

    1 2 3 4 5 5 6 7 8

    States

    0000

    ComplexityRelated to the stabilization periodEstimation : replication or ergodic estimation

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 11 / 27

  • Classical forward simulation

    ForwardRepresentation : transition fonction

    Xn+1 = Φ(Xn, en+1).

    x ← x0{choice of the initial state at time =0}n = 0;repeat

    n ← n + 1;e ← Random event();x ← Φ(x, e);{computation of the next state Xn+1}

    until some empirical criteriareturn x

    Convergence : biased sampleSampling : Warm-up period

    Trajectory

    Initial state

    0001

    0010

    0011

    0100

    0101

    0110

    1000

    1001

    1010

    1100

    0

    Time

    1 2 3 4 5 5 6 7 8

    States

    0000

    ComplexityRelated to the stabilization periodEstimation : replication or ergodic estimation

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 11 / 27

  • Classical forward simulation

    ForwardRepresentation : transition fonction

    Xn+1 = Φ(Xn, en+1).

    x ← x0{choice of the initial state at time =0}n = 0;repeat

    n ← n + 1;e ← Random event();x ← Φ(x, e);{computation of the next state Xn+1}

    until some empirical criteriareturn x

    Convergence : biased sampleSampling : Warm-up period

    Trajectory

    Initial state

    0001

    0010

    0011

    0100

    0101

    0110

    1000

    1001

    1010

    1100

    0

    Time

    1 2 3 4 5 5 6 7 8

    States

    0000

    ComplexityRelated to the stabilization periodEstimation : replication or ergodic estimation

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 11 / 27

  • Classical forward simulation

    ForwardRepresentation : transition fonction

    Xn+1 = Φ(Xn, en+1).

    x ← x0{choice of the initial state at time =0}n = 0;repeat

    n ← n + 1;e ← Random event();x ← Φ(x, e);{computation of the next state Xn+1}

    until some empirical criteriareturn x

    Convergence : biased sampleSampling : Warm-up period

    Trajectory

    Initial state

    0001

    0010

    0011

    0100

    0101

    0110

    1000

    1001

    1010

    1100

    0

    Time

    1 2 3 4 5 5 6 7 8

    States

    0000

    ComplexityRelated to the stabilization periodEstimation : replication or ergodic estimation

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 11 / 27

  • Classical forward simulation

    ForwardRepresentation : transition fonction

    Xn+1 = Φ(Xn, en+1).

    x ← x0{choice of the initial state at time =0}n = 0;repeat

    n ← n + 1;e ← Random event();x ← Φ(x, e);{computation of the next state Xn+1}

    until some empirical criteriareturn x

    Convergence : biased sampleSampling : Warm-up period

    Trajectory

    6

    0010

    0011

    0100

    0101

    0110

    1000

    1001

    1010

    1100

    0

    Initial state

    Time

    1 2 3 4 5 7 8 9

    States

    0000

    0001

    ComplexityRelated to the stabilization periodEstimation : replication or ergodic estimation

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 11 / 27

  • Classical forward simulation

    ForwardRepresentation : transition fonction

    Xn+1 = Φ(Xn, en+1).

    x ← x0{choice of the initial state at time =0}n = 0;repeat

    n ← n + 1;e ← Random event();x ← Φ(x, e);{computation of the next state Xn+1}

    until some empirical criteriareturn x

    Convergence : biased sampleSampling : Warm-up period

    Trajectory

    6

    0010

    0011

    0100

    0101

    0110

    1000

    1001

    1010

    1100

    0

    Initial state

    Time

    1 2 3 4 5 7 8 9

    States

    0000

    0001

    ComplexityRelated to the stabilization periodEstimation : replication or ergodic estimation

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 11 / 27

  • Classical forward simulation

    ForwardRepresentation : transition fonction

    Xn+1 = Φ(Xn, en+1).

    x ← x0{choice of the initial state at time =0}n = 0;repeat

    n ← n + 1;e ← Random event();x ← Φ(x, e);{computation of the next state Xn+1}

    until some empirical criteriareturn x

    Convergence : biased sampleSampling : Warm-up period

    Trajectory

    6

    0010

    0011

    0100

    0101

    0110

    1000

    1001

    1010

    1100

    0

    Initial state

    Time

    1 2 3 4 5 7 8 9

    States

    0000

    0001

    ComplexityRelated to the stabilization periodEstimation : replication or ergodic estimation

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 11 / 27

  • Classical forward simulation

    ForwardRepresentation : transition fonction

    Xn+1 = Φ(Xn, en+1).

    x ← x0{choice of the initial state at time =0}n = 0;repeat

    n ← n + 1;e ← Random event();x ← Φ(x, e);{computation of the next state Xn+1}

    until some empirical criteriareturn x

    Convergence : biased sampleSampling : Warm-up period

    Trajectory

    6

    0010

    0011

    0100

    0101

    0110

    1000

    1001

    1010

    1100

    0

    Initial state

    Time

    1 2 3 4 5 7 8 9

    States

    0000

    0001

    ComplexityRelated to the stabilization periodEstimation : replication or ergodic estimation

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 11 / 27

  • Classical forward simulation

    ForwardRepresentation : transition fonction

    Xn+1 = Φ(Xn, en+1).

    x ← x0{choice of the initial state at time =0}n = 0;repeat

    n ← n + 1;e ← Random event();x ← Φ(x, e);{computation of the next state Xn+1}

    until some empirical criteriareturn x

    Convergence : biased sampleSampling : Warm-up period

    Trajectory

    stabilization period

    0101

    0110

    1000

    1001

    1010

    1100

    0

    Initial state

    6

    Steady

    state ?

    Time

    1 2 3 4 5 7 8 9

    States

    0000

    0001

    0010

    0011

    0100

    ComplexityRelated to the stabilization periodEstimation : replication or ergodic estimation

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 11 / 27

  • Classical forward simulation

    ForwardRepresentation : transition fonction

    Xn+1 = Φ(Xn, en+1).

    x ← x0{choice of the initial state at time =0}n = 0;repeat

    n ← n + 1;e ← Random event();x ← Φ(x, e);{computation of the next state Xn+1}

    until some empirical criteriareturn x

    Convergence : biased sampleSampling : Warm-up period

    Trajectory

    stabilization period

    0101

    0110

    1000

    1001

    1010

    1100

    0

    Initial state

    6

    Steady

    state ?

    Time

    1 2 3 4 5 7 8 9

    States

    0000

    0001

    0010

    0011

    0100

    ComplexityRelated to the stabilization periodEstimation : replication or ergodic estimation

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 11 / 27

  • Perfect simulation : backward idea

    Representation : transition fonction

    Xn+1 = Φ(Xn, en+1), {en}n∈Z i.i.d. sequence.In what state could I be at time n = 0 ?

    X0 ∈ X = Z0∈ Φ(X , e0) = Z1∈ Φ(Φ(X , e−1), e0) = Z2· · ·∈ Φ(Φ(· · ·Φ(X , e−n+1), · · · ), e0) = Zn

    TheoremProvided some condition on the sequence of events, the sequence of sets

    Z0 ⊇ Z1 ⊇ Z2 ⊇ · · · ⊇ Zn ⊇ · · · is decreasing to a single state.

    The generated state is stationary distributed (steady state sample).

    τb = inf{n ∈ N; Card(Zn) = 1}.

    backward coupling time

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 12 / 27

  • Perfect simulation : backward idea

    Representation : transition fonction

    Xn+1 = Φ(Xn, en+1), {en}n∈Z i.i.d. sequence.In what state could I be at time n = 0 ?

    X0 ∈ X = Z0∈ Φ(X , e0) = Z1∈ Φ(Φ(X , e−1), e0) = Z2· · ·∈ Φ(Φ(· · ·Φ(X , e−n+1), · · · ), e0) = Zn

    TheoremProvided some condition on the sequence of events, the sequence of sets

    Z0 ⊇ Z1 ⊇ Z2 ⊇ · · · ⊇ Zn ⊇ · · · is decreasing to a single state.

    The generated state is stationary distributed (steady state sample).

    τb = inf{n ∈ N; Card(Zn) = 1}.

    backward coupling time

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 12 / 27

  • Perfect simulation

    Backward algorithmRepresentation : transition fonction

    Xn+1 = Φ(Xn, en+1).

    for all x ∈ X doy(x) ← x

    end forrepeat

    u ← Random;for all x ∈ X do

    e ← Random event();y(x) ← y(Φ(x, e));

    end foruntil All y(x) are equalreturn y(x)

    Convergence : If the algorithm stops, thereturned value is steady state distributedCoupling time: τ < +∞, properties of Φ

    Trajectories

    Mean time complexitycΦ mean computation cost of Φ(x , e)

    C 6 Card(X ).Eτ.cΦ.J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 13 / 27

  • Perfect simulation

    Backward algorithmRepresentation : transition fonction

    Xn+1 = Φ(Xn, en+1).

    for all x ∈ X doy(x) ← x

    end forrepeat

    u ← Random;for all x ∈ X do

    e ← Random event();y(x) ← y(Φ(x, e));

    end foruntil All y(x) are equalreturn y(x)

    Convergence : If the algorithm stops, thereturned value is steady state distributedCoupling time: τ < +∞, properties of Φ

    Trajectories

    Mean time complexitycΦ mean computation cost of Φ(x , e)

    C 6 Card(X ).Eτ.cΦ.J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 13 / 27

  • Perfect simulation

    Backward algorithmRepresentation : transition fonction

    Xn+1 = Φ(Xn, en+1).

    for all x ∈ X doy(x) ← x

    end forrepeat

    u ← Random;for all x ∈ X do

    e ← Random event();y(x) ← y(Φ(x, e));

    end foruntil All y(x) are equalreturn y(x)

    Convergence : If the algorithm stops, thereturned value is steady state distributedCoupling time: τ < +∞, properties of Φ

    Trajectories

    Time

    States

    0000

    0001

    0010

    0011

    0100

    0101

    0110

    1000

    1001

    1010

    1100

    −4 −3 −2 −1−5−6−7−8−9−10 0

    Mean time complexitycΦ mean computation cost of Φ(x , e)

    C 6 Card(X ).Eτ.cΦ.J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 13 / 27

  • Perfect simulation

    Backward algorithmRepresentation : transition fonction

    Xn+1 = Φ(Xn, en+1).

    for all x ∈ X doy(x) ← x

    end forrepeat

    u ← Random;for all x ∈ X do

    e ← Random event();y(x) ← y(Φ(x, e));

    end foruntil All y(x) are equalreturn y(x)

    Convergence : If the algorithm stops, thereturned value is steady state distributedCoupling time: τ < +∞, properties of Φ

    Trajectories

    Time

    States

    0000

    0001

    0010

    0011

    0100

    0101

    0110

    1000

    1001

    1010

    1100

    −4 −3 −2 −1−5−6−7−8−9−10 0U1

    Mean time complexitycΦ mean computation cost of Φ(x , e)

    C 6 Card(X ).Eτ.cΦ.J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 13 / 27

  • Perfect simulation

    Backward algorithmRepresentation : transition fonction

    Xn+1 = Φ(Xn, en+1).

    for all x ∈ X doy(x) ← x

    end forrepeat

    u ← Random;for all x ∈ X do

    e ← Random event();y(x) ← y(Φ(x, e));

    end foruntil All y(x) are equalreturn y(x)

    Convergence : If the algorithm stops, thereturned value is steady state distributedCoupling time: τ < +∞, properties of Φ

    Trajectories

    Time

    States

    0000

    0001

    0010

    0011

    0100

    0101

    0110

    1000

    1001

    1010

    1100

    −4 −3 −2 −1−5−6−7−8−9−10 0U1

    Mean time complexitycΦ mean computation cost of Φ(x , e)

    C 6 Card(X ).Eτ.cΦ.J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 13 / 27

  • Perfect simulation

    Backward algorithmRepresentation : transition fonction

    Xn+1 = Φ(Xn, en+1).

    for all x ∈ X doy(x) ← x

    end forrepeat

    u ← Random;for all x ∈ X do

    e ← Random event();y(x) ← y(Φ(x, e));

    end foruntil All y(x) are equalreturn y(x)

    Convergence : If the algorithm stops, thereturned value is steady state distributedCoupling time: τ < +∞, properties of Φ

    Trajectories

    Time

    States

    0000

    0001

    0010

    0011

    0100

    0101

    0110

    1000

    1001

    1010

    1100

    −4 −3 −2 −1−5−6−7−8−9−10 0U1U2

    Mean time complexitycΦ mean computation cost of Φ(x , e)

    C 6 Card(X ).Eτ.cΦ.J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 13 / 27

  • Perfect simulation

    Backward algorithmRepresentation : transition fonction

    Xn+1 = Φ(Xn, en+1).

    for all x ∈ X doy(x) ← x

    end forrepeat

    u ← Random;for all x ∈ X do

    e ← Random event();y(x) ← y(Φ(x, e));

    end foruntil All y(x) are equalreturn y(x)

    Convergence : If the algorithm stops, thereturned value is steady state distributedCoupling time: τ < +∞, properties of Φ

    Trajectories

    Time

    States

    0000

    0001

    0010

    0011

    0100

    0101

    0110

    1000

    1001

    1010

    1100

    −4 −3 −2 −1−5−6−7−8−9−10 0U1U2

    Mean time complexitycΦ mean computation cost of Φ(x , e)

    C 6 Card(X ).Eτ.cΦ.J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 13 / 27

  • Perfect simulation

    Backward algorithmRepresentation : transition fonction

    Xn+1 = Φ(Xn, en+1).

    for all x ∈ X doy(x) ← x

    end forrepeat

    u ← Random;for all x ∈ X do

    e ← Random event();y(x) ← y(Φ(x, e));

    end foruntil All y(x) are equalreturn y(x)

    Convergence : If the algorithm stops, thereturned value is steady state distributedCoupling time: τ < +∞, properties of Φ

    Trajectories

    Time

    States

    0000

    0001

    0010

    0011

    0100

    0101

    0110

    1000

    1001

    1010

    1100

    −4 −3 −2 −1−5−6−7−8−9−10 0U1U2U3

    Mean time complexitycΦ mean computation cost of Φ(x , e)

    C 6 Card(X ).Eτ.cΦ.J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 13 / 27

  • Perfect simulation

    Backward algorithmRepresentation : transition fonction

    Xn+1 = Φ(Xn, en+1).

    for all x ∈ X doy(x) ← x

    end forrepeat

    u ← Random;for all x ∈ X do

    e ← Random event();y(x) ← y(Φ(x, e));

    end foruntil All y(x) are equalreturn y(x)

    Convergence : If the algorithm stops, thereturned value is steady state distributedCoupling time: τ < +∞, properties of Φ

    Trajectories

    Time

    States

    0000

    0001

    0010

    0011

    0100

    0101

    0110

    1000

    1001

    1010

    1100

    −4 −3 −2 −1−5−6−7−8−9−10 0U1U2U3

    Mean time complexitycΦ mean computation cost of Φ(x , e)

    C 6 Card(X ).Eτ.cΦ.J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 13 / 27

  • Perfect simulation

    Backward algorithmRepresentation : transition fonction

    Xn+1 = Φ(Xn, en+1).

    for all x ∈ X doy(x) ← x

    end forrepeat

    u ← Random;for all x ∈ X do

    e ← Random event();y(x) ← y(Φ(x, e));

    end foruntil All y(x) are equalreturn y(x)

    Convergence : If the algorithm stops, thereturned value is steady state distributedCoupling time: τ < +∞, properties of Φ

    Trajectories

    Time

    States

    0000

    0001

    0010

    0011

    0100

    0101

    0110

    1000

    1001

    1010

    1100

    −4 −3 −2 −1−5−6−7−8−9−10 0U1U2U3U4

    Mean time complexitycΦ mean computation cost of Φ(x , e)

    C 6 Card(X ).Eτ.cΦ.J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 13 / 27

  • Perfect simulation

    Backward algorithmRepresentation : transition fonction

    Xn+1 = Φ(Xn, en+1).

    for all x ∈ X doy(x) ← x

    end forrepeat

    u ← Random;for all x ∈ X do

    e ← Random event();y(x) ← y(Φ(x, e));

    end foruntil All y(x) are equalreturn y(x)

    Convergence : If the algorithm stops, thereturned value is steady state distributedCoupling time: τ < +∞, properties of Φ

    Trajectories

    Time

    States

    0000

    0001

    0010

    0011

    0100

    0101

    0110

    1000

    1001

    1010

    1100

    −4 −3 −2 −1−5−6−7−8−9−10 0U1U2U3U4

    Mean time complexitycΦ mean computation cost of Φ(x , e)

    C 6 Card(X ).Eτ.cΦ.J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 13 / 27

  • Perfect simulation

    Backward algorithmRepresentation : transition fonction

    Xn+1 = Φ(Xn, en+1).

    for all x ∈ X doy(x) ← x

    end forrepeat

    u ← Random;for all x ∈ X do

    e ← Random event();y(x) ← y(Φ(x, e));

    end foruntil All y(x) are equalreturn y(x)

    Convergence : If the algorithm stops, thereturned value is steady state distributedCoupling time: τ < +∞, properties of Φ

    Trajectories

    Time

    States

    0000

    0001

    0010

    0011

    0100

    0101

    0110

    1000

    1001

    1010

    1100

    −4 −3 −2 −1−5−6−7−8−9−10 0U1U2U3U4U5

    Mean time complexitycΦ mean computation cost of Φ(x , e)

    C 6 Card(X ).Eτ.cΦ.J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 13 / 27

  • Perfect simulation

    Backward algorithmRepresentation : transition fonction

    Xn+1 = Φ(Xn, en+1).

    for all x ∈ X doy(x) ← x

    end forrepeat

    u ← Random;for all x ∈ X do

    e ← Random event();y(x) ← y(Φ(x, e));

    end foruntil All y(x) are equalreturn y(x)

    Convergence : If the algorithm stops, thereturned value is steady state distributedCoupling time: τ < +∞, properties of Φ

    Trajectories

    Time

    States

    0000

    0001

    0010

    0011

    0100

    0101

    0110

    1000

    1001

    1010

    1100

    −4 −3 −2 −1−5−6−7−8−9−10 0U1U2U3U4U5

    Mean time complexitycΦ mean computation cost of Φ(x , e)

    C 6 Card(X ).Eτ.cΦ.J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 13 / 27

  • Perfect simulation

    Backward algorithmRepresentation : transition fonction

    Xn+1 = Φ(Xn, en+1).

    for all x ∈ X doy(x) ← x

    end forrepeat

    u ← Random;for all x ∈ X do

    e ← Random event();y(x) ← y(Φ(x, e));

    end foruntil All y(x) are equalreturn y(x)

    Convergence : If the algorithm stops, thereturned value is steady state distributedCoupling time: τ < +∞, properties of Φ

    Trajectories

    Time

    States

    0000

    0001

    0010

    0011

    0100

    0101

    0110

    1000

    1001

    1010

    1100

    −4 −3 −2 −1−5−6−7−8−9−10 0U1U2U3U4U5U6

    Mean time complexitycΦ mean computation cost of Φ(x , e)

    C 6 Card(X ).Eτ.cΦ.J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 13 / 27

  • Perfect simulation

    Backward algorithmRepresentation : transition fonction

    Xn+1 = Φ(Xn, en+1).

    for all x ∈ X doy(x) ← x

    end forrepeat

    u ← Random;for all x ∈ X do

    e ← Random event();y(x) ← y(Φ(x, e));

    end foruntil All y(x) are equalreturn y(x)

    Convergence : If the algorithm stops, thereturned value is steady state distributedCoupling time: τ < +∞, properties of Φ

    Trajectories

    Time

    States

    0000

    0001

    0010

    0011

    0100

    0101

    0110

    1000

    1001

    1010

    1100

    −4 −3 −2 −1−5−6−7−8−9−10 0U1U2U3U4U5U6

    Mean time complexitycΦ mean computation cost of Φ(x , e)

    C 6 Card(X ).Eτ.cΦ.J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 13 / 27

  • Perfect simulation

    Backward algorithmRepresentation : transition fonction

    Xn+1 = Φ(Xn, en+1).

    for all x ∈ X doy(x) ← x

    end forrepeat

    u ← Random;for all x ∈ X do

    e ← Random event();y(x) ← y(Φ(x, e));

    end foruntil All y(x) are equalreturn y(x)

    Convergence : If the algorithm stops, thereturned value is steady state distributedCoupling time: τ < +∞, properties of Φ

    Trajectories

    Time

    States

    0000

    0001

    0010

    0011

    0100

    0101

    0110

    1000

    1001

    1010

    1100

    −4 −3 −2 −1−5−6−7−8−9−10 0U1U2U3U4U5U6U7

    Mean time complexitycΦ mean computation cost of Φ(x , e)

    C 6 Card(X ).Eτ.cΦ.J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 13 / 27

  • Perfect simulation

    Backward algorithmRepresentation : transition fonction

    Xn+1 = Φ(Xn, en+1).

    for all x ∈ X doy(x) ← x

    end forrepeat

    u ← Random;for all x ∈ X do

    e ← Random event();y(x) ← y(Φ(x, e));

    end foruntil All y(x) are equalreturn y(x)

    Convergence : If the algorithm stops, thereturned value is steady state distributedCoupling time: τ < +∞, properties of Φ

    Trajectories

    Time

    States

    0000

    0001

    0010

    0011

    0100

    0101

    0110

    1000

    1001

    1010

    1100

    −4 −3 −2 −1−5−6−7−8−9−10 0U1U2U3U4U5U6U7

    Mean time complexitycΦ mean computation cost of Φ(x , e)

    C 6 Card(X ).Eτ.cΦ.J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 13 / 27

  • Perfect simulation

    Backward algorithmRepresentation : transition fonction

    Xn+1 = Φ(Xn, en+1).

    for all x ∈ X doy(x) ← x

    end forrepeat

    u ← Random;for all x ∈ X do

    e ← Random event();y(x) ← y(Φ(x, e));

    end foruntil All y(x) are equalreturn y(x)

    Convergence : If the algorithm stops, thereturned value is steady state distributedCoupling time: τ < +∞, properties of Φ

    Trajectories

    Time

    States

    0000

    0001

    0010

    0011

    0100

    0101

    0110

    1000

    1001

    1010

    1100

    −4 −3 −2 −1−5−6−7−8−9−10 0U1U2U3U4U5U6U7U8

    Mean time complexitycΦ mean computation cost of Φ(x , e)

    C 6 Card(X ).Eτ.cΦ.J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 13 / 27

  • Perfect simulation

    Backward algorithmRepresentation : transition fonction

    Xn+1 = Φ(Xn, en+1).

    for all x ∈ X doy(x) ← x

    end forrepeat

    u ← Random;for all x ∈ X do

    e ← Random event();y(x) ← y(Φ(x, e));

    end foruntil All y(x) are equalreturn y(x)

    Convergence : If the algorithm stops, thereturned value is steady state distributedCoupling time: τ < +∞, properties of Φ

    Trajectories

    Time

    States

    0000

    0001

    0010

    0011

    0100

    0101

    0110

    1000

    1001

    1010

    1100

    −4 −3 −2 −1−5−6−7−8−9−10 0U1U2U3U4U5U6U7U8

    Mean time complexitycΦ mean computation cost of Φ(x , e)

    C 6 Card(X ).Eτ.cΦ.J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 13 / 27

  • Perfect simulation

    Backward algorithmRepresentation : transition fonction

    Xn+1 = Φ(Xn, en+1).

    for all x ∈ X doy(x) ← x

    end forrepeat

    u ← Random;for all x ∈ X do

    e ← Random event();y(x) ← y(Φ(x, e));

    end foruntil All y(x) are equalreturn y(x)

    Convergence : If the algorithm stops, thereturned value is steady state distributedCoupling time: τ < +∞, properties of Φ

    Trajectories

    Time

    States

    0000

    0001

    0010

    0011

    0100

    0101

    0110

    1000

    1001

    1010

    1100

    −4 −3 −2 −1−5−6−7−8−9−10 0U1U2U3U4U5U6U7U8

    τ∗

    Mean time complexitycΦ mean computation cost of Φ(x , e)

    C 6 Card(X ).Eτ.cΦ.J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 13 / 27

  • Perfect simulation

    Backward algorithmRepresentation : transition fonction

    Xn+1 = Φ(Xn, en+1).

    for all x ∈ X doy(x) ← x

    end forrepeat

    u ← Random;for all x ∈ X do

    e ← Random event();y(x) ← y(Φ(x, e));

    end foruntil All y(x) are equalreturn y(x)

    Convergence : If the algorithm stops, thereturned value is steady state distributedCoupling time: τ < +∞, properties of Φ

    Trajectories

    Time

    States

    0000

    0001

    0010

    0011

    0100

    0101

    0110

    1000

    1001

    1010

    1100

    −4 −3 −2 −1−5−6−7−8−9−10 0U1U2U3U4U5U6U7U8

    τ∗

    Mean time complexitycΦ mean computation cost of Φ(x , e)

    C 6 Card(X ).Eτ.cΦ.J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 13 / 27

  • Monotonicity and perfect simulation : idea

    min = (0, · · · , 0) and Max = (C1, · · · , Cn).

    If all events are monotone then

    X0 ∈ Zn ⊂ [Φ(min, e−n→0),Φ(Max , e−n→0)]

    ⇒ 2 trajectories

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 14 / 27

  • Monotonicity and perfect simulation

    Monotone PSDoubling scheme

    n=1;R[1]=Random event;repeat

    n=2.n;y(min) ← miny(Max) ← Maxfor i=n downto n/2+1 do

    R[i]=Random event;end forfor i=n downto 1 do

    y(min) ← Φ(y(min), R[i])y(Max) ← Φ(y(Max), R[i])

    end foruntil y(min) = y(Max)return y(min)

    Trajectories

    : minimum

    2

    1

    M

    −1−2−4−8−16−32 0

    States

    : : Maximum

    0

    Mean time complexity

    Cm 6 2.(2.Eτ).cΦ. Reduction factor : 4Card(X ) .

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 15 / 27

  • Monotonicity and perfect simulation

    Monotone PSDoubling scheme

    n=1;R[1]=Random event;repeat

    n=2.n;y(min) ← miny(Max) ← Maxfor i=n downto n/2+1 do

    R[i]=Random event;end forfor i=n downto 1 do

    y(min) ← Φ(y(min), R[i])y(Max) ← Φ(y(Max), R[i])

    end foruntil y(min) = y(Max)return y(min)

    Trajectories

    : minimum

    2

    1

    M

    −1−2−4−8−16−32 0

    States

    : : Maximum

    0

    Mean time complexity

    Cm 6 2.(2.Eτ).cΦ. Reduction factor : 4Card(X ) .

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 15 / 27

  • Monotonicity and perfect simulation

    Monotone PSDoubling scheme

    n=1;R[1]=Random event;repeat

    n=2.n;y(min) ← miny(Max) ← Maxfor i=n downto n/2+1 do

    R[i]=Random event;end forfor i=n downto 1 do

    y(min) ← Φ(y(min), R[i])y(Max) ← Φ(y(Max), R[i])

    end foruntil y(min) = y(Max)return y(min)

    Trajectories

    : minimum

    2

    1

    M

    −1−2−4−8−16−32 0

    States

    : : Maximum

    0

    Mean time complexity

    Cm 6 2.(2.Eτ).cΦ. Reduction factor : 4Card(X ) .

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 15 / 27

  • Monotonicity and perfect simulation

    Monotone PSDoubling scheme

    n=1;R[1]=Random event;repeat

    n=2.n;y(min) ← miny(Max) ← Maxfor i=n downto n/2+1 do

    R[i]=Random event;end forfor i=n downto 1 do

    y(min) ← Φ(y(min), R[i])y(Max) ← Φ(y(Max), R[i])

    end foruntil y(min) = y(Max)return y(min)

    Trajectories

    : minimum

    2

    1

    M

    −1−2−4−8−16−32 0

    States

    : : Maximum

    0

    Mean time complexity

    Cm 6 2.(2.Eτ).cΦ. Reduction factor : 4Card(X ) .

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 15 / 27

  • Monotonicity and perfect simulation

    Monotone PSDoubling scheme

    n=1;R[1]=Random event;repeat

    n=2.n;y(min) ← miny(Max) ← Maxfor i=n downto n/2+1 do

    R[i]=Random event;end forfor i=n downto 1 do

    y(min) ← Φ(y(min), R[i])y(Max) ← Φ(y(Max), R[i])

    end foruntil y(min) = y(Max)return y(min)

    Trajectories

    : minimum

    2

    1

    M

    −1−2−4−8−16−32 0

    States

    : : Maximum

    0

    Mean time complexity

    Cm 6 2.(2.Eτ).cΦ. Reduction factor : 4Card(X ) .

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 15 / 27

  • Monotonicity and perfect simulation

    Monotone PSDoubling scheme

    n=1;R[1]=Random event;repeat

    n=2.n;y(min) ← miny(Max) ← Maxfor i=n downto n/2+1 do

    R[i]=Random event;end forfor i=n downto 1 do

    y(min) ← Φ(y(min), R[i])y(Max) ← Φ(y(Max), R[i])

    end foruntil y(min) = y(Max)return y(min)

    Trajectories

    : minimum

    2

    1

    M

    −1−2−4−8−16−32 0

    States

    : : Maximum

    0

    Mean time complexity

    Cm 6 2.(2.Eτ).cΦ. Reduction factor : 4Card(X ) .

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 15 / 27

  • Monotonicity and perfect simulation

    Monotone PSDoubling scheme

    n=1;R[1]=Random event;repeat

    n=2.n;y(min) ← miny(Max) ← Maxfor i=n downto n/2+1 do

    R[i]=Random event;end forfor i=n downto 1 do

    y(min) ← Φ(y(min), R[i])y(Max) ← Φ(y(Max), R[i])

    end foruntil y(min) = y(Max)return y(min)

    Trajectories

    : minimum

    2

    1

    M

    −1−2−4−8−16−32 0

    States

    : : Maximum

    0

    Mean time complexity

    Cm 6 2.(2.Eτ).cΦ. Reduction factor : 4Card(X ) .

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 15 / 27

  • Monotonicity and perfect simulation

    Monotone PSDoubling scheme

    n=1;R[1]=Random event;repeat

    n=2.n;y(min) ← miny(Max) ← Maxfor i=n downto n/2+1 do

    R[i]=Random event;end forfor i=n downto 1 do

    y(min) ← Φ(y(min), R[i])y(Max) ← Φ(y(Max), R[i])

    end foruntil y(min) = y(Max)return y(min)

    Trajectories

    : minimum

    2

    1

    M

    −1−2−4−8−16−32 0

    States

    : : Maximum

    0

    Mean time complexity

    Cm 6 2.(2.Eτ).cΦ. Reduction factor : 4Card(X ) .

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 15 / 27

  • Monotonicity and perfect simulation

    Monotone PSDoubling scheme

    n=1;R[1]=Random event;repeat

    n=2.n;y(min) ← miny(Max) ← Maxfor i=n downto n/2+1 do

    R[i]=Random event;end forfor i=n downto 1 do

    y(min) ← Φ(y(min), R[i])y(Max) ← Φ(y(Max), R[i])

    end foruntil y(min) = y(Max)return y(min)

    Trajectories

    State

    2

    1

    M

    −1−2−4−8−16−32 0

    States

    : : Maximum

    : minimum

    Generated

    0

    Mean time complexity

    Cm 6 2.(2.Eτ).cΦ. Reduction factor : 4Card(X ) .

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 15 / 27

  • Monotonicity and perfect simulation

    Monotone PSDoubling scheme

    n=1;R[1]=Random event;repeat

    n=2.n;y(min) ← miny(Max) ← Maxfor i=n downto n/2+1 do

    R[i]=Random event;end forfor i=n downto 1 do

    y(min) ← Φ(y(min), R[i])y(Max) ← Φ(y(Max), R[i])

    end foruntil y(min) = y(Max)return y(min)

    Trajectories

    State

    2

    1

    M

    −1−2−4−8−16−32 0

    States

    : : Maximum

    : minimum

    Generated

    0

    Mean time complexity

    Cm 6 2.(2.Eτ).cΦ. Reduction factor : 4Card(X ) .

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 15 / 27

  • Coupling time

    definition

    τb = min{n ∈ N; Card(Zn) = 1};= min{n ∈ N; |Φ(X , e−n→0| = 1}.

    Properties

    - Backward τb and forward τ f coupling times have the sameprobability distribution;

    - Marginal coupling : denote by τbi the backward coupling time forQi

    τb = max τbi .

    Problem : compute the mean coupling time

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 16 / 27

  • Coupling time

    definition

    τb = min{n ∈ N; Card(Zn) = 1};= min{n ∈ N; |Φ(X , e−n→0| = 1}.

    Properties

    - Backward τb and forward τ f coupling times have the sameprobability distribution;

    - Marginal coupling : denote by τbi the backward coupling time forQi

    τb = max τbi .

    Problem : compute the mean coupling time

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 16 / 27

  • Coupling time

    definition

    τb = min{n ∈ N; Card(Zn) = 1};= min{n ∈ N; |Φ(X , e−n→0| = 1}.

    Properties

    - Backward τb and forward τ f coupling times have the sameprobability distribution;

    - Marginal coupling : denote by τbi the backward coupling time forQi

    τb = max τbi .

    Problem : compute the mean coupling time

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 16 / 27

  • Outline

    1 Queueing Networks with finite capacity

    2 Event modelling and monotonicity

    3 Perfect simulation and coupling time

    4 Acyclic networks

    5 Synthesis and future works

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 17 / 27

  • Coupling experiment

    Queueing model :

    5

    C

    C

    C

    C0

    1

    2

    λ

    λ

    λ

    λλ

    01

    2

    3

    4

    Estimation of Eτ :

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 18 / 27

  • Coupling experiment

    Queueing model :

    5

    C

    C

    C

    C0

    1

    2

    λ

    λ

    λ

    λλ

    01

    2

    3

    4

    Estimation of Eτ :

    0

    50

    100

    150

    200

    250

    300

    350

    400

    τ

    0 1 2 3 4 λ

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 18 / 27

  • Main result

    Theorem (Bound on coupling time)

    Eτ 6K∑

    i=1

    Λ

    Λi

    Ci + C2i2

    ,

    - Λ : global event rate in the network,

    - Λi the rate of events affecting Qi- Ci is the capacity of Queue i.

    Sketch of the proof- Explicit computation for the M/M/1/C

    - Computable bounds for the M/M/1/C

    - Bound with isolated queues

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 19 / 27

  • Main result

    Theorem (Bound on coupling time)

    Eτ 6K∑

    i=1

    Λ

    Λi

    Ci + C2i2

    ,

    - Λ : global event rate in the network,

    - Λi the rate of events affecting Qi- Ci is the capacity of Queue i.

    Sketch of the proof- Explicit computation for the M/M/1/C

    - Computable bounds for the M/M/1/C

    - Bound with isolated queues

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 19 / 27

  • Explicit computation for the M/M/1/C

    Eτb = E min(h0→C , hC→0)Absorbing time in a finite Markov chain; p = λλ+µ = 1− q

    1,C

    1,C−10,C−2

    1,C−2 2,C−1

    2,C

    3,C

    C−2,C−1 C−1,C

    C,C0,0

    0,1 1,2

    0,C

    0,C−1

    0,C−3

    p

    p

    p

    p

    p

    pp p p p

    ppp

    p p

    pq

    q

    q

    q

    q

    q q q q q

    qqq

    q q

    qLevel 3

    Level 4

    Level 5

    Level C+1

    Level C+2

    Level 2

    Explicit recurrence equationsCase λ = µ Eτb = C+C22 .

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 20 / 27

  • Computable bounds for M/M/1/C

    If the stationary distribution is concentrated on 0 (λ < µ),

    Eτb 6 Eh0→C is an accurate bound.

    Theorem

    The mean coupling time Eτb of a M/M/1/C queue with arrival rate λand service rate µ is bounded using p = λ/(λ + µ) = 1− q.

    Critical bound: ∀p ∈ [0, 1], Eτb 6 C2+C2 .

    Heavy traffic Bound: if p > 12 , Eτb 6 Cp−q −

    q(1−“

    qp

    ”C)

    (p−q)2.

    Light traffic bound: if p < 12 , Eτb 6 Cq−p −

    p(1−“

    pq

    ”C)

    (q−p)2.

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 21 / 27

  • Computable bounds for M/M/1/C

    Example with C = 10

    0

    20

    40

    60

    80

    100

    120

    0 0.2 0.4 0.6 0.8 1

    Eτb

    p

    heavy trafficLight trafficbound

    C+C2

    2

    C + C2

    bound

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 22 / 27

  • Example for tandem queues

    Coupling of Queue 0

    Time

    0

    X 00 = 55

    4

    2

    1

    3 = C1

    6 = C0

    0

    −τ b0Coupling of queue 1 conditionned by state of queue 0

    4

    3 = C1

    1

    0

    −τ b1 (s0 = 2) 0Time

    X 11 = 2

    X 10 = 3

    5 = X 00

    6

    2

    X 11 = 2

    5

    4

    2

    1

    3 = C1

    6 = C0

    0

    −τ b0 − τ b1 (s0 = 5)

    X 00 = 5

    τ b1 (s0 = 5) 0Time

    X 10 = 3

    Then τb 6st ∞τb1 + τb0 , normalized

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 23 / 27

  • Bound with isolated queues

    TheoremIn an acyclic stable network of K M/M/1/Ci queues with Bernoullirouting and losses in case of overflow, the coupling time from the pastsatisfies in expectation,

    E[τb] 6K−1∑i=0

    Λ

    `i + µi

    Ciqi − pi

    −pi(1−

    (piqi

    )Ci)

    (qi − pi)2

    6

    K−1∑i=0

    Λ

    `i + µi(Ci + C

    2i ).

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 24 / 27

  • Outline

    1 Queueing Networks with finite capacity

    2 Event modelling and monotonicity

    3 Perfect simulation and coupling time

    4 Acyclic networks

    5 Synthesis and future works

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 25 / 27

  • Synthesis

    Computable bound for the mean coupling time :

    - linear in the number of component of the model;

    - at most quadratic in queues sizes;

    - large capacity queues ( bound is accurate).

    Practical impact

    - Accurate bounds, dimensionning of trajectories length;

    - Simulation useful even for low probability events;

    - Coupling time is explained by the spread of the stationarydistribution.

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 26 / 27

  • Future works

    Conjecture for general networks.

    0

    700

    800

    0 0.5 1 1.5 2 2.5 3 3.5 4

    500

    400

    300

    200

    100

    600

    λ5

    Eτb

    B1 (proven)

    B1 ∧ B2 ∧ B3

    B3 (conjecture)

    B2 (conjecture)

    Extension to cyclic networks,Generalization to several types of eventsApplication : Grid and call centers

    J.G. Dopper, B. Gaujal and J.-M. Vincent (Universities of Grenoble)Bounds for the Coupling Time in Queueing Networks Perfect SimulationMAM 2006, june12 27 / 27

    Queueing Networks with finite capacityEvent modelling and monotonicityPerfect simulation and coupling timeAcyclic networksSynthesis and future works