IEEE ICCA 2010 – Xiamen, June 11, 2010 On Closed Form Solutions for Equilibrium Probabilities in the Closed Lu-Kumar Network under Various Buffer Priority.

Slides:



Advertisements
Similar presentations
Lecture 10 Queueing Theory. There are a few basic elements common to almost all queueing theory application. Customers arrive, they wait for service in.
Advertisements

Modeling and Dimensioning of Mobile Networks: from GSM to LTE
Lecture 6  Calculating P n – how do we raise a matrix to the n th power?  Ergodicity in Markov Chains.  When does a chain have equilibrium probabilities?
1 A class of Generalized Stochastic Petri Nets for the performance Evaluation of Mulitprocessor Systems By M. Almone, G. Conte Presented by Yinglei Song.
©2013 – James R. Morrison – UIUC ISE Seminar – August 22, 2013 Recent Directions in the Theory of Flow Lines with Applications to Semiconductor Manufacturing.
TCOM 501: Networking Theory & Fundamentals
Copyright © 2005 Department of Computer Science CPSC 641 Winter Markov Chains Plan: –Introduce basics of Markov models –Define terminology for Markov.
Continuous Time Markov Chains and Basic Queueing Theory
NETE4631:Capacity Planning (3)- Private Cloud Lecture 11 Suronapee Phoomvuthisarn, Ph.D. /
Queueing Theory: Recap
Oct State-Space Modeling Slide 1 Fault-Tolerant Computing Motivation, Background, and Tools.
CSE 3504: Probabilistic Analysis of Computer Systems Topics covered: Continuous time Markov chains (Sec )
CSE 221: Probabilistic Analysis of Computer Systems Topics covered: Simple queuing models (Sec )
TCOM 501: Networking Theory & Fundamentals
Mean Delay in M/G/1 Queues with Head-of-Line Priority Service and Embedded Markov Chains Wade Trappe.
1 Performance Evaluation of Computer Networks Objectives  Introduction to Queuing Theory  Little’s Theorem  Standard Notation of Queuing Systems  Poisson.
Single queue modeling. Basic definitions for performance predictions The performance of a system that gives services could be seen from two different.
2003 Fall Queuing Theory Midterm Exam(Time limit:2 hours)
Modeling and Analysis of Manufacturing Systems Session 2 QUEUEING MODELS January 2001.
1 TCOM 501: Networking Theory & Fundamentals Lectures 9 & 10 M/G/1 Queue Prof. Yannis A. Korilis.
Queuing Networks: Burke’s Theorem, Kleinrock’s Approximation, and Jackson’s Theorem Wade Trappe.
1 TCOM 501: Networking Theory & Fundamentals Lecture 8 March 19, 2003 Prof. Yannis A. Korilis.
1 Markov Chains H Plan: –Introduce basics of Markov models –Define terminology for Markov chains –Discuss properties of Markov chains –Show examples of.
Lecture 14 – Queuing Systems
Dimitrios Konstantas, Evangelos Grigoroudis, Vassilis S. Kouikoglou and Stratos Ioannidis Department of Production Engineering and Management Technical.
Flows and Networks Plan for today (lecture 5): Last time / Questions? Waiting time simple queue Little Sojourn time tandem network Jackson network: mean.
An Analysis of Chaining Protocols for Video-on-Demand J.-F. Pâris University of Houston Thomas Schwarz, S. J. Universidad Católica del Uruguay.
Introduction to Discrete Event Simulation Customer population Service system Served customers Waiting line Priority rule Service facilities Figure C.1.
Flows and Networks Plan for today (lecture 5): Last time / Questions? Blocking of transitions Kelly / Whittle network Optimal design of a Kelly / Whittle.
©2009 – James R. Morrison – IEEE CASE 2009 – August 25, 2009 Regular Flow Line Models for Semiconductor Cluster Tools: James R. Morrison Assistant Professor.
MIT Fun queues for MIT The importance of queues When do queues appear? –Systems in which some serving entities provide some service in a shared.
Lecture 10: Queueing Theory. Queueing Analysis Jobs serviced by the system resources Jobs wait in a queue to use a busy server queueserver.
Flows and Networks Plan for today (lecture 6): Last time / Questions? Kelly / Whittle network Optimal design of a Kelly / Whittle network: optimisation.
Analysis of M/M/c/N Queuing System With Balking, Reneging and Synchronous Vacations Dequan Yue Department of Statistics, College of Sciences Yanshan University,
Introduction to Queueing Theory
Networks of Queues Plan for today (lecture 6): Last time / Questions? Product form preserving blocking Interpretation traffic equations Kelly / Whittle.
Flows and Networks Plan for today (lecture 4): Last time / Questions? Output simple queue Tandem network Jackson network: definition Jackson network: equilibrium.
TexPoint fonts used in EMF.
1 Elements of Queuing Theory The queuing model –Core components; –Notation; –Parameters and performance measures –Characteristics; Markov Process –Discrete-time.
CDA6530: Performance Models of Computers and Networks Chapter 7: Basic Queuing Networks TexPoint fonts used in EMF. Read the TexPoint manual before you.
Markov Chains X(t) is a Markov Process if, for arbitrary times t1 < t2 < < tk < tk+1 If X(t) is discrete-valued If X(t) is continuous-valued i.e.
Model under consideration: Loss system Collection of resources to which calls with holding time  (c) and class c arrive at random instances. An arriving.
1 Markov chains and processes: motivations Random walk One-dimensional walk You can only move one step right or left every time unit Two-dimensional walk.
NETE4631: Network Information System Capacity Planning (2) Suronapee Phoomvuthisarn, Ph.D. /
XS3D Lab Seminar, September 10, 2014, Sang Yoon Bae – 1 xS3D Summer Research Focus Period Lab Seminar Presentation: Sang Yoon Bae Department of Industrial.
Chapter 6 Product-Form Queuing Network Models Prof. Ali Movaghar.
Chapter 5 Elementary Stochastic Analysis Prof. Ali Movaghar.
© 2015 McGraw-Hill Education. All rights reserved. Chapter 17 Queueing Theory.
Flows and Networks Plan for today (lecture 3): Last time / Questions? Output simple queue Tandem network Jackson network: definition Jackson network: equilibrium.
Flows and Networks Plan for today (lecture 6): Last time / Questions? Kelly / Whittle network Optimal design of a Kelly / Whittle network: optimisation.
QUEUING. CONTINUOUS TIME MARKOV CHAINS {X(t), t >= 0} is a continuous time process with > sojourn times S 0, S 1, S 2,... > embedded process X n = X(S.
Networks (3TU) Plan for today (lecture 5): Last time / Questions? Tandem network Jackson network: definition Jackson network: equilibrium distribution.
Discrete-time Markov chain (DTMC) State space distribution
Industrial Engineering Dep
Much More About Markov Chains
Department of Industrial Engineering
Flows and Networks Plan for today (lecture 4):
DTMC Applications Ranking Web Pages & Slotted ALOHA
Internet Queuing Delay Introduction
Lecture on Markov Chain
Finite M/M/1 queue Consider an M/M/1 queue with finite waiting room.
IV-2 Manufacturing Systems modeling
Markov Chains Carey Williamson Department of Computer Science
Lecture 4: Algorithmic Methods for G/M/1 and M/G/1 type models
Queueing networks.
Carey Williamson Department of Computer Science University of Calgary
Queuing Theory III.
Queuing Theory III.
Discrete-time markov chain (continuation)
Erlang, Hyper-exponential, and Coxian distributions
Presentation transcript:

IEEE ICCA 2010 – Xiamen, June 11, 2010 On Closed Form Solutions for Equilibrium Probabilities in the Closed Lu-Kumar Network under Various Buffer Priority Policies Seunghwan Jung and James R. Morrison KAIST, Department of Industrial and Systems Engineering IEEE ICCA 2010 Xiamen, China June 11, 2010

IEEE ICCA 2010 – Xiamen, June 11, Presentation Overview Introduction System Description Equilibrium Probabilities Under the LBFS Policy Equilibrium Probabilities Under the FBFS Policy Conclusion

IEEE ICCA 2010 – Xiamen, June 11, Introduction Server 1Server 2 Customers arrive Customers exit Jackson network is one of the rare class of network that possess closed form equilibrium probability distributions.

IEEE ICCA 2010 – Xiamen, June 11, Introduction Except for some classes of networks, few networks possess closed form equilibrium probability distributions. [1] James R. Morrison, “Implementation of a Fluctuation Smoothing Production Control Policy in IBM’s 200mm Wafer Fab”, European Control Conference, pp , 2005.

IEEE ICCA 2010 – Xiamen, June 11, Introduction Obtain closed form equilibrium probabilities. Allows complete characterization of the steady state behavior.

IEEE ICCA 2010 – Xiamen, June 11, System Description: Network Model Two stations : σ 1 and σ 2 Buffers : b 1, b 2, b 3, b 4 Service time for a customer in buffer b i : exponential with rate μ i N trapped customers circulate within the network A closed reentrant queueing network

IEEE ICCA 2010 – Xiamen, June 11, System Description: Last Buffer First Served Non-idling, preemptive Gives priority b 1 over b 4 and b 3 over b 2 A closed reentrant queueing network

IEEE ICCA 2010 – Xiamen, June 11, System Description: First Buffer First Served Non-idling, preemptive Gives priority b 4 over b 1 and b 2 over b 3 A closed reentrant queueing network

IEEE ICCA 2010 – Xiamen, June 11, Equilibrium Probabilities under LBFS System state at time t : S(t)={w(t),x(t),y(t),z(t)} w(t),x(t),y(t),z(t) : Number of customers in buffers b 1, b 2, b 3, b 4 at time t Uniformization : Get Discrete time Markov chain Steady state probability of state S : Π s A closed reentrant queueing network Transition diagram under LBFS 1 N-1 00

IEEE ICCA 2010 – Xiamen, June 11, Equilibrium Probabilities under LBFS Transition diagram under LBFS To find equilibrium probability : Balance equations Π = Π P “Flow in” = “Flow out” So, assuming that we know, we can obtain. So we can express in terms of Recursively, we can express whole steady state probabilities in terms of initial condition.

IEEE ICCA 2010 – Xiamen, June 11, Equilibrium Probabilities under LBFS To specify our main idea, we redefine the state as below :

IEEE ICCA 2010 – Xiamen, June 11, Equilibrium Probabilities under LBFS Overall steps for obtaining closed form solutions Step 1: We make the equation involving only one type of signal by combining given equations Step 2: Taking z-transform and inverting it give a closed form solution for the signal Step 3: Plugging the closed form solution into the other balance equations gives closed form solutions for them

IEEE ICCA 2010 – Xiamen, June 11, Equilibrium Probabilities under LBFS Overall steps for obtaining closed form solutions (continued) Step 4: Using the balance equations, all X k [n] are expressed in terms of X 0 [0] Step 5: Summing all probabilities and setting them equal to 1 to get X 0 [0]

IEEE ICCA 2010 – Xiamen, June 11, Equilibrium Probabilities under FBFS A closed reentrant queueing network Transition diagram under FBFS System state at time t : S(t)={w(t),x(t),y(t),z(t)} w(t),x(t),y(t),z(t) : Number of customers in buffers b 1, b 2, b 3, b 4 at time t Uniformization : Get Discrete time Markov chain Steady state probability of state S : Π s

IEEE ICCA 2010 – Xiamen, June 11, Equilibrium Probabilities under FBFS Transition diagram under FBFS Recursively, we can express whole steady state probabilities in terms of initial conditions. To find equilibrium probability : Balance equations Π = Π P “Flow in” = “Flow out” Initial conditions So, assuming that we know, we can obtain.

IEEE ICCA 2010 – Xiamen, June 11, Equilibrium Probabilities under FBFS To specify our main idea, we redefine the state as below :

IEEE ICCA 2010 – Xiamen, June 11, Equilibrium Probabilities under FBFS Overall steps for obtaining closed form solutions Step 1: Investigating X 0 [n], we obtain relationship below: Step 2: Using relationship between X k [m] and X k-1 [n], we obtain X 1 [n]. Step 3: Recursively, we can obtain

IEEE ICCA 2010 – Xiamen, June 11, Equilibrium Probabilities under FBFS Step 4: By symmetry, we get the inverse transforms for the lower region Step 5: Using remaining balance equations, we express all X k [n] in terms of X 0 [0].(Toeplitz matrix structure)

IEEE ICCA 2010 – Xiamen, June 11, Equilibrium Probabilities under FBFS Step 5: Summing all probabilities and setting them equal to 1 to get X 0 [0] Note: Not a complete closed form

IEEE ICCA 2010 – Xiamen, June 11, Concluding Remarks LBFS : Indeed obtained a closed form solution FBFS : Enough structure to reduce the computational complexity To obtain equilibrium probabilities by “Π = Π P ”, we have to inverse (N+1) 2 ╳ (N+1) 2 matrix. Future works Attempting to obtain a closed-form expression for the inverse of the Toeplitz matrix from the FBFS case. Extend the structure to more general cases.