Delay bounds (Simon S. Lam) 1
|
|
- Peregrine George
- 6 years ago
- Views:
Transcription
1 1 Pacet Scheduling: End-to-End E d Delay Bounds Delay bounds (Simon S. Lam) 1
2 2 Reerences Delay Guarantee o Virtual Cloc server Georey G. Xie and Simon S. Lam, Delay Guarantee o Virtual Cloc Server, IEEE/ACM Transactions on Networing, Vol. 3, No. 6, December End-to-End E Delay Guarantees and Bounds Simon S. Lam and Georey G. Xie, Group Priority Scheduling, IEEE/ACM Transactions on Networing, Vol. Vo. 5, No. 2, April Pawan Goyal, Simon S. Lam, and Harric Vin, Determining End-to-End E Delay Bounds in Heterogeneous Networs, ACM/Springer-Verlag Multimedia Systems, Vol. 5, No. 3, May Delay bounds (Simon S. Lam) 2
3 3 Virtual Cloc (VC) server - review Let r be the reserved rate in bits/s allocated to low. Let p denote a pacet in low, with length l(p) bits and arrival time, A(p) ( 0) Each low has a virtual cloc, priority(), which h is zero initially and updated d whenever a new pacet in low arrives: l( p) priority ( ) max{ priority ( ), A( p)} + (1) r The new value o priority() ) is assigned to pacet p as its virtual cloc value, denoted by P(p) Delay bounds (Simon S. Lam) 3
4 4 VC server review (cont.) Whenever the VC server is ready or another pacet, the pacet among all lows with the smallest virtual cloc value is selected FCFS within a low non-preemptive Meaning o virtual cloc value Consider a hypothetical server with rate r dedicated to low (processor sharing). Let F(p) denote the inishing time o pacet p. In this system, l ( p + 1) F ( p + 1) = max{ F( p), A( p + 1)} + r Virtual cloc value P(p) in previous slide is same as inishing time F(p) in processor sharing Delay bounds (Simon S. Lam) 4
5 Delay Guarantee o VC server VC by itsel does not provide a delay bound in the usual sense, i.e., L(p) A(p) cannot be bounded by the server alone, where L(p) denotes the departure time o pacet p Given admission control, VC can provide a delay guarantee to a low with no assumption that sources are low- controlled or well behaved It is a conditional guarantee based upon a low s reserved rate Delay bounds (Simon S. Lam) 5 5
6 Active lows VC server or a set F o lows, with reserved rate r allocated to low, or in F Deinition 1: A low is active at time t i priority () > t (2) That is, a low is active i its virtual cloc is running aster than real time. The virtual cloc o low is driven by the low s arrivals and pacet lengths only [see eq. (1)] Delay bounds (Simon S. Lam) 6 6
7 7 Active low interpretation At time t, the condition priority() > t holds at the hypothetical ti server (processor-sharing) o low i and only i the hypothetical server is baclogged i.e., there is at least one low pacet waiting or being served During a busy period o the VC server (pacet by pacet), low may be active when there is no low pacet in the system (queue + server); low may be inactive when the system has low pacets Delay bounds (Simon S. Lam) 7
8 8 Capacity constraint condition Deintion 2: Let C denote the capacity, in bits/s, o a VC server. The server s capacity is not exceeded at time t i the ollowing condition holds: a () t r C (3) where a(t), a subset o F, is the set o lows that are active at time t Above condition using a(t) allows more pacets to be admitted than using F but requires admission control or each pacet arrival Delay bounds (Simon S. Lam) 8
9 9 Delay guarantee theorem Theorem 1: I the capacity o a VC server has not been exceeded or a non-zero duration since the start o a busy period, then the ollowing holds or every pacet p served during the busy period: lmax L ( p ) P ( p ) + (4) C where l is the maximum pacet size max (Proo by contradiction) This bound is analogous to Pareh and Gallager s bound relating PGPS to GPS Delay bounds (Simon S. Lam) 9
10 10 Observations abouttheorem 1 The theorem holds without any assumption that sources are well-behaved or lowcontrolled. While each low, say, has a reserved rate r, its source can generate pacets at a much larger rate than r. Pacets can have arbitrary arrival times and pacet lengths The delay guarantee in (4) holds even i the lows in F generate pacets at an aggregate rate larger than C [assuming each low is allocated its own buers] It is a conditional guarantee based on the low s reserved rate Delay bounds (Simon S. Lam) 10
11 11 VC server needs to enorce the capacity constraint t The set o active lows, a(t), changes dynamically. At any time the server can determine whether low is active by comparing priority() with the current time t. Thus the server can determine the set a(t) In theory, the server may perorm admission control upon the arrival o each pacet to prevent violation o its capacity constraint Delay bounds (Simon S. Lam) 11
12 12 Practical admission controls Without a priori nowledge o source arrival characteristics: Flow sources are statically ti assigned reserved rates. Thus or all time t r r C a() t F On-demand assignment per session. A source can generate traic only ater it has been assigned a reserved rate on demand: r r C a() t D() t where D(t) is the set o sessions ss at time t with assigned reserved rates Delay bounds (Simon S. Lam) 12
13 13 Practical admission controls (more) On-demand assignment per application data unit (burst) All pacets o an application data unit are admitted or discarded together Examples: (i) a picture in a video, (ii) a chun o a video Given a priori nowledge o source traic statistics, a server can over-commit its capacity, maing use o the act that only some lows are active at any time Statistical rather than deterministic guarantee Delay bounds (Simon S. Lam) 13
14 14 Properties o VC delay guarantee A delay guarantee o the orm L(p) P(p) + β, where β is a constant, is a conditional guarantee A pacet s delay is bounded rom its expected arrival time based upon its reserved rate, that is, arriving early does not help On the other hand, pacets arriving late do not get better service or not using their low s reserved rate This encourages on-time arrivals, which are much better or buer management in a switch (router) Delay bounds (Simon S. Lam) 14
15 15 Delay Bounds Consider a low with reserved rate r and pacets indexed by n = 1, 2, in order o arrival. I P(n)-A(n) is bounded above, then the delay o pacet n, L(n)-A(n), is bounded above The goal o source control is to bound P(n) A(n) Method 1: Ensure that the inter-arrival time o two consecutive pacets is bounded below, ln ( ) ln ( ) An ( + 1) An ( ) Pn ( ) = An ( ) + r r max ( ) Ln ( ) Pn ( ) + l Thus, Ln ( ) An ( ) l n + l C r C max Delay bounds (Simon S. Lam) 15
16 16 Leay Bucet source control Let Arrival ( t1, t2) denote the number o low bits that arrive at server in interval [t 1, t 2 ] Bits o a pacet arrive when the last bit o the pacet arrives The arrival unction consists o a jump at each pacet arrival time and is right continuous, Arrival ( t, t) is the size o the pacet that arrives at time t Flow conorms to Leay Bucet with bucet size σ and rate ρ i and only i Arrival ( t, t ) σ + ρ ( t t ) or 0 t t Delay bounds (Simon S. Lam) 16
17 17 Delay Bounds (cont.) ( σ, ρ) Method 2: Source control by Leay Bucet with bucet size σ and toen arrival rate ρ equal to the reserved rate o the low, r It is proved [Goyal, Lam, Vin 1997] that or all n P ( n ) A ( n ) σ ρ σ Thus, we have Ln ( ) An ( ) + r For Spring semester 2017, sip ahead to slide 40 l max C Delay bounds (Simon S. Lam) 17
18 18 Generalizing Theorem 1 The delay guarantee theorem holds as long as the server capacity is not exceeded by the sum o the reserved rates o lows that t have had a pacet served during the busy period. We can change the active low deinition (to admit more pacets) Deinition 1 : A low is active at time t i (i) the system is not empty, (ii) a low pacet has arrived in the current busy period, (iii) priority() > t Delay bounds (Simon S. Lam) 18
19 19 Generalizing Theorem 1 by another way We can achieve the generalization, without changing the active low deinition, by resetting the virtual cloc o each low to zero at the end o a busy period. Speciically, when the system becomes empty upon a pacet departure, then or all in F, priority() <- 0 Side eect: Resetting means that the debt o low is orgotten (orgiven) Good or bad? Delay bounds (Simon S. Lam) 19
20 20 Deterministic End-to-End Delay Guarantees and Bounds Delay bounds (Simon S. Lam) 20
21 21 End-to-End Delay Guarantee Delay bounds (Simon S. Lam) 21
22 22 Notation or server Delay bounds (Simon S. Lam) 22
23 23 Guaranteed Deadline (GD) Server It provides the ollowing service to low e.g., a Virtual Cloc server Delay bounds (Simon S. Lam) 23
24 24 Traversing a path o GD servers = β + τ,+1 Delay bounds (Simon S. Lam) 24
25 25 Notation or low Delay bounds (Simon S. Lam) 25
26 26 Reerence cloc values or low v V () i () i V a time constant associated with pacet i (seconds) reerence cloc value o pacet i at server, to be interpreted as expected inishing time o pacet i at server (0) is 0 by deinition and or i 1 V ( i) = max{ V ( i 1), A ( i)}+ v ( i) (3) For the special case o VC servers v ( i) = s ( i) / λ ( i) where λ ( i) is reserved rate or pacet i in low The virtual cloc value o pacet i at server is its reerence cloc value and or all i P () i = V () i Delay bounds (Simon S. Lam) 26
27 27 Important relations Each server ensures that pacet i in low departs by its deadline which is P () i + β A ( i) depends on P ( i) as shown in (2) + 1 V ( i) depends on A ( i) as shown in (3) For a VC server,, we now that P () i = V () i What about other scheduling disciplines? Delay bounds (Simon S. Lam) 27
28 28 Putting the relations together Consider pacet i in low. Its end to end delay is A () i V () i 1 1 A () 1 i A () i A () i K σ ρ <- source control A () i () A + 1 i K () A i A () 1 i K K+1 source L () i L L + i () 1 () i + 1 LK () i P () i + β A () i = L () i + τ P () i + α + 1, + 1 β K + destination <- deadline guarantee by server is delay due to non-preemption Delay bounds (Simon S. Lam) 28
29 29 Dierence between a pacet s deadline and reerence cloc value For each GD server, suppose we now, or all i, P () i V () i Proo in [Lam and Xie 1997] by induction Recall that v s ( j) ( j) = and λ ( j) α = β + α +, 1 Delay bounds (Simon S. Lam) 29
30 30 End-to-end delay guarantee theorem Proo in [Lam and Xie 1997] by applying Lemma 2 recursively Delay bounds (Simon S. Lam) 30
31 31 End-to-end delay upper bound AK ( j ) A ( j ) The end-to-end delay o pacet j is K The end-to-end guarantee in (5) provides an upper bound on the end-to-end delay i a source control mechanism is used such that and V ( j) A ( j) 1 1 has a inite upper bound, server is a GD server, 1 K, such that the term P ( j ) V ( j ) has a inite upper bound Delay bounds (Simon S. Lam) 31
32 32 Example o source control I the source o low is controlled by a leay bucet with bucet depth σ and toen rate ρ = λ, which is the reserved rate o low, then or all pacet i in the low [Goyal, Lam, Vin 1997] To obtain an end-to-end delay upper bound or low, is instantiated to in the delay guarantee ormula o Theorem 1. Delay bounds (Simon S. Lam) 32
33 33 End-to-end delay upper bound (cont.) Note that dierent service disciplines may be chosen or dierent servers along a path also, the term P ( j) V ( j) may be positive or negative With Theorem 1, the problem o providing an upper bound on the end-to-end ddelay or a low along a path o heterogenous servers is reduced to a set o singlenode problems! Delay bounds (Simon S. Lam) 33
34 34 Examples o GD servers The GD class o servers is general including many pacet schedulers in the literature. They dier in their P (.) unctions, β constants, and admission control conditions We will show our: Virtual Cloc WFQ/PGPS Delay-EDD Leave-in-Time Delay bounds (Simon S. Lam) 34
35 35 1. Virtual Cloc (VC) as a GD server Delay bounds (Simon S. Lam) 35
36 36 2. WFQ/PGPS as a GD server The pacet deadline at server is, using P ( j) = L, GPS( j), s L ( j) L ( j ) rom Pareh and Gallager age [1993] max, PGPS j, GPS j + C where L ( j) is the departure time o being a GPS server GPS, { } with weights w g, g F At GPS server, consider the rate, λ where b ( t) is the set o baclogged lows, to be the reserved rate or low wc g w g b () t Set o lows must be inite or the reserved rate to be nonzero or all t Delay bounds (Simon S. Lam) 36
37 37 2..WFQ/PGPS as a GD server (cont.) From Goyal, Lam, and Vin [1997] eq. (7) s ( j ) LGPS, ( j) max{ LGPS, ( j 1), A ( j)} +. λ V ( j) = max{ V ( j 1), A ( j)} + v (j) rom eq. (3) s ( j) where v ( j)=. λ Thereore, P ( j) = L ( j) V ( j), GPS Finally, P ( j) V ( j) 0 Delay bounds (Simon S. Lam) 37
38 38 3. Delay-EDD as a GD server > 0 Delay bounds (Simon S. Lam) 38
39 39 4. Leave-in-Time as a GD server > 0 Delay bounds (Simon S. Lam) 39
40 40 End-to-end delay bound or a path o servers sthat tuse VC or WFQ/PGPS scheduling Using Eq. (5) o Theorem 1 together with leay bucet source control, we get σ A i A i K ( j) K K+ 1() 1 () + ( 1)max + 1 j i ρ ρ =1= 1 σ + ( K 1)max 1 j is ( j) K α ρ = 1 = +, + 1 s α A tight bound where ρ λ and α = β + τ, where β = s max C Delay bounds (Simon S. Lam) 40
41 41 The end-to-end delay bound o Pareh and Gallager A. Pareh and R. Gallager, A generalized processor-sharing approach to low control in integrated services networs: the multiple node case, IEEE/ACM Trans Networing, April 1994, equation (39). Previously nown bound or rate-proportional processor sharing (RPPS) rate assignment o PGPS networs when the sources conorm to Leay Bucet is σ + 2( K 1) s K max ρ = 1 max + α a loose bound where s is the maximum pacet size or low Delay bounds (Simon S. Lam) 41
42 42 Summary End-to-end delay bounds can be provided to lows by a variety o GD servers Delay guarantee theorem Decomposition o the end-to-end delay bound problem into single-node and source control problems Both VC and WFQ/PGPS servers provide the best delay bound (a tight bound) VC is easier to implement WFQ is slightly more air Delay bounds (Simon S. Lam) 42
43 43 The end Delay bounds (Simon S. Lam) 43
Max-Weight Scheduling in Queueing Networks with. Heavy-Tailed Traffic
Max-Weight Scheduling in Queueing Networks with Heavy-Tailed Traic arxiv:1108.0370v1 [cs.ni] 1 Aug 011 Mihalis G. Markakis, Eytan H. Modiano, and John N. Tsitsiklis Abstract We consider the problem o packet
More informationIntro to Queueing Theory
1 Intro to Queueing Theory Little s Law M/G/1 queue Conservation Law 1/31/017 M/G/1 queue (Simon S. Lam) 1 Little s Law No assumptions applicable to any system whose arrivals and departures are observable
More informationComplexity and Algorithms for Two-Stage Flexible Flowshop Scheduling with Availability Constraints
Complexity and Algorithms or Two-Stage Flexible Flowshop Scheduling with Availability Constraints Jinxing Xie, Xijun Wang Department o Mathematical Sciences, Tsinghua University, Beijing 100084, China
More informationCyclic Schedules: General Structure. Frame Size Constraints
CPSC-663: Real-ime Systems Cyclic Schedules: General Structure Scheduling decision is made periodically: Frame Scheduling decision is made periodically: choose which job to execute perorm monitoring and
More informationSome properties of variable length packet shapers
Some properties of variable length packet shapers Jean-Yves Le Boudec EPFL-DSC CH-1015 Lausanne, Switzerland EPFL-DSC esearch eport DSC/2000/037 20 Decembre 2000 Abstract The
More informationReal-time operating systems course. 6 Definitions Non real-time scheduling algorithms Real-time scheduling algorithm
Real-time operating systems course 6 Definitions Non real-time scheduling algorithms Real-time scheduling algorithm Definitions Scheduling Scheduling is the activity of selecting which process/thread should
More informationA Generalized Processor Sharing Approach to Flow Control in Integrated Services Networks: The Single Node Case. 1
A Generalized Processor Sharing Approach to Flow Control in Integrated Services Networks: The Single Node Case 1 Abhay K Parekh 2 3 and Robert G Gallager 4 Laboratory for Information and Decision Systems
More informationM/G/FQ: STOCHASTIC ANALYSIS OF FAIR QUEUEING SYSTEMS
M/G/FQ: STOCHASTIC ANALYSIS OF FAIR QUEUEING SYSTEMS MOHAMMED HAWA AND DAVID W. PETR Information and Telecommunications Technology Center University of Kansas, Lawrence, Kansas, 66045 email: {hawa, dwp}@ittc.ku.edu
More informationLecture 6. Real-Time Systems. Dynamic Priority Scheduling
Real-Time Systems Lecture 6 Dynamic Priority Scheduling Online scheduling with dynamic priorities: Earliest Deadline First scheduling CPU utilization bound Optimality and comparison with RM: Schedulability
More information5/15/18. Operations Research: An Introduction Hamdy A. Taha. Copyright 2011, 2007 by Pearson Education, Inc. All rights reserved.
The objective of queuing analysis is to offer a reasonably satisfactory service to waiting customers. Unlike the other tools of OR, queuing theory is not an optimization technique. Rather, it determines
More informationTDDB68 Concurrent programming and operating systems. Lecture: CPU Scheduling II
TDDB68 Concurrent programming and operating systems Lecture: CPU Scheduling II Mikael Asplund, Senior Lecturer Real-time Systems Laboratory Department of Computer and Information Science Copyright Notice:
More informationExtending the Network Calculus Pay Bursts Only Once Principle to Aggregate Scheduling
Extending the Networ Calculus Pay Bursts Only Once Principle to Aggregate Scheduling Marus Fidler Department of Computer Science, Aachen University Ahornstr. 55, 52074 Aachen, Germany fidler@i4.informati.rwth-aachen.de
More informationProblem Set. Problems on Unordered Summation. Math 5323, Fall Februray 15, 2001 ANSWERS
Problem Set Problems on Unordered Summation Math 5323, Fall 2001 Februray 15, 2001 ANSWERS i 1 Unordered Sums o Real Terms In calculus and real analysis, one deines the convergence o an ininite series
More informationBounding the End-to-End Response Times of Tasks in a Distributed. Real-Time System Using the Direct Synchronization Protocol.
Bounding the End-to-End Response imes of asks in a Distributed Real-ime System Using the Direct Synchronization Protocol Jun Sun Jane Liu Abstract In a distributed real-time system, a task may consist
More informationModule 5: CPU Scheduling
Module 5: CPU Scheduling Basic Concepts Scheduling Criteria Scheduling Algorithms Multiple-Processor Scheduling Real-Time Scheduling Algorithm Evaluation 5.1 Basic Concepts Maximum CPU utilization obtained
More informationChapter 6: CPU Scheduling
Chapter 6: CPU Scheduling Basic Concepts Scheduling Criteria Scheduling Algorithms Multiple-Processor Scheduling Real-Time Scheduling Algorithm Evaluation 6.1 Basic Concepts Maximum CPU utilization obtained
More informationClock-driven scheduling
Clock-driven scheduling Also known as static or off-line scheduling Michal Sojka Czech Technical University in Prague, Faculty of Electrical Engineering, Department of Control Engineering November 8, 2017
More informationTardiness Bounds under Global EDF Scheduling on a Multiprocessor
Tardiness ounds under Global EDF Scheduling on a Multiprocessor UmaMaheswari C. Devi and James H. Anderson Department of Computer Science The University of North Carolina at Chapel Hill Abstract This paper
More informationEE 368. Weeks 3 (Notes)
EE 368 Weeks 3 (Notes) 1 State of a Queuing System State: Set of parameters that describe the condition of the system at a point in time. Why do we need it? Average size of Queue Average waiting time How
More informationAperiodic Task Scheduling
Aperiodic Task Scheduling Jian-Jia Chen (slides are based on Peter Marwedel) TU Dortmund, Informatik 12 Germany Springer, 2010 2017 年 11 月 29 日 These slides use Microsoft clip arts. Microsoft copyright
More informationADMISSION AND FLOW CONTROL IN GENERALIZED PROCESSOR SHARING SCHEDULERS
ADMISSION AND FLOW CONTROL IN GENERALIZED PROCESSOR SHARING SCHEDULERS Ph.D. Theses By Róbert Szabó Research Supervisors: Dr. Tibor Trón Dr. József Bíró Department of Telecommunications and Telematics
More informationTraffic models on a network of roads
Traic models on a network o roads Alberto Bressan Department o Mathematics, Penn State University bressan@math.psu.edu Center or Interdisciplinary Mathematics Alberto Bressan (Penn State) Traic low on
More information1.225 Transportation Flow Systems Quiz (December 17, 2001; Duration: 3 hours)
1.225 Transportation Flow Systems Quiz (December 17, 2001; Duration: 3 hours) Student Name: Alias: Instructions: 1. This exam is open-book 2. No cooperation is permitted 3. Please write down your name
More informationA Measurement-Analytic Approach for QoS Estimation in a Network Based on the Dominant Time Scale
222 IEEE/ACM TRANSACTIONS ON NETWORKING, VOL. 11, NO. 2, APRIL 2003 A Measurement-Analytic Approach for QoS Estimation in a Network Based on the Dominant Time Scale Do Young Eun and Ness B. Shroff, Senior
More informationCS 361 Meeting 28 11/14/18
CS 361 Meeting 28 11/14/18 Announcements 1. Homework 9 due Friday Computation Histories 1. Some very interesting proos o undecidability rely on the technique o constructing a language that describes the
More informationTask Models and Scheduling
Task Models and Scheduling Jan Reineke Saarland University June 27 th, 2013 With thanks to Jian-Jia Chen at KIT! Jan Reineke Task Models and Scheduling June 27 th, 2013 1 / 36 Task Models and Scheduling
More informationAdvanced Computer Networks Lecture 3. Models of Queuing
Advanced Computer Networks Lecture 3. Models of Queuing Husheng Li Min Kao Department of Electrical Engineering and Computer Science University of Tennessee, Knoxville Spring, 2016 1/13 Terminology of
More informationReal-Time Systems. Event-Driven Scheduling
Real-Time Systems Event-Driven Scheduling Marcus Völp, Hermann Härtig WS 2013/14 Outline mostly following Jane Liu, Real-Time Systems Principles Scheduling EDF and LST as dynamic scheduling methods Fixed
More informationDefinition: Let f(x) be a function of one variable with continuous derivatives of all orders at a the point x 0, then the series.
2.4 Local properties o unctions o several variables In this section we will learn how to address three kinds o problems which are o great importance in the ield o applied mathematics: how to obtain the
More informationReal-Time Systems. Event-Driven Scheduling
Real-Time Systems Event-Driven Scheduling Hermann Härtig WS 2018/19 Outline mostly following Jane Liu, Real-Time Systems Principles Scheduling EDF and LST as dynamic scheduling methods Fixed Priority schedulers
More informationNon-Preemptive and Limited Preemptive Scheduling. LS 12, TU Dortmund
Non-Preemptive and Limited Preemptive Scheduling LS 12, TU Dortmund 09 May 2017 (LS 12, TU Dortmund) 1 / 31 Outline Non-Preemptive Scheduling A General View Exact Schedulability Test Pessimistic Schedulability
More information[4] T. I. Seidman, \\First Come First Serve" is Unstable!," tech. rep., University of Maryland Baltimore County, 1993.
[2] C. J. Chase and P. J. Ramadge, \On real-time scheduling policies for exible manufacturing systems," IEEE Trans. Automat. Control, vol. AC-37, pp. 491{496, April 1992. [3] S. H. Lu and P. R. Kumar,
More informationSolutions to COMP9334 Week 8 Sample Problems
Solutions to COMP9334 Week 8 Sample Problems Problem 1: Customers arrive at a grocery store s checkout counter according to a Poisson process with rate 1 per minute. Each customer carries a number of items
More informationScheduling Coflows in Datacenter Networks: Improved Bound for Total Weighted Completion Time
1 1 2 Scheduling Coflows in Datacenter Networs: Improved Bound for Total Weighted Completion Time Mehrnoosh Shafiee and Javad Ghaderi Abstract Coflow is a recently proposed networing abstraction to capture
More informationA New Approach for Asynchronous Distributed Rate Control of Elastic Sessions in Integrated Packet Networks
A New Approach or Asynchronous Distributed Rate Control o Elastic Sessions in Integrated Packet Networks Santosh P. Abraham and Anurag Kumar Abstract We develop a new class o asynchronous distributed algorithms
More informationControlling Burstiness in Fair Queueing Scheduling
TEL-AVIV UNIVERSITY RAYMOND AND BEVERLY SACKLER FACULTY OF EXACT SCIENCES SCHOOL OF COMPUTER SCIENCES Controlling Burstiness in Fair Queueing Scheduling Thesis submitted in partial fulfillment of the requirements
More informationQueueing Theory and Simulation. Introduction
Queueing Theory and Simulation Based on the slides of Dr. Dharma P. Agrawal, University of Cincinnati and Dr. Hiroyuki Ohsaki Graduate School of Information Science & Technology, Osaka University, Japan
More informationCaching Policy Optimization for Video on Demand
Caching Policy Optimization or Video on Demand Hao Wang, Shengqian Han, and Chenyang Yang School o Electronics and Inormation Engineering, Beihang University, China, 100191 Email: {wang.hao, sqhan, cyyang}@buaa.edu.cn
More informationDesign of IP networks with Quality of Service
Course of Multimedia Internet (Sub-course Reti Internet Multimediali ), AA 2010-2011 Prof. Pag. 1 Design of IP networks with Quality of Service 1 Course of Multimedia Internet (Sub-course Reti Internet
More informationEmbedded Systems 15. REVIEW: Aperiodic scheduling. C i J i 0 a i s i f i d i
Embedded Systems 15-1 - REVIEW: Aperiodic scheduling C i J i 0 a i s i f i d i Given: A set of non-periodic tasks {J 1,, J n } with arrival times a i, deadlines d i, computation times C i precedence constraints
More informationNon-preemptive Fixed Priority Scheduling of Hard Real-Time Periodic Tasks
Non-preemptive Fixed Priority Scheduling of Hard Real-Time Periodic Tasks Moonju Park Ubiquitous Computing Lab., IBM Korea, Seoul, Korea mjupark@kr.ibm.com Abstract. This paper addresses the problem of
More informationPaper Presentation. Amo Guangmo Tong. University of Taxes at Dallas January 24, 2014
Paper Presentation Amo Guangmo Tong University of Taxes at Dallas gxt140030@utdallas.edu January 24, 2014 Amo Guangmo Tong (UTD) January 24, 2014 1 / 30 Overview 1 Tardiness Bounds under Global EDF Scheduling
More informationIEEE TRANSACTIONS ON AUTOMATIC CONTROL, VOL. 43, NO. 3, MARCH
IEEE TRANSACTIONS ON AUTOMATIC CONTROL, VOL. 43, NO. 3, MARCH 1998 315 Asymptotic Buffer Overflow Probabilities in Multiclass Multiplexers: An Optimal Control Approach Dimitris Bertsimas, Ioannis Ch. Paschalidis,
More informationOn-line scheduling of periodic tasks in RT OS
On-line scheduling of periodic tasks in RT OS Even if RT OS is used, it is needed to set up the task priority. The scheduling problem is solved on two levels: fixed priority assignment by RMS dynamic scheduling
More informationRecent Progress on a Statistical Network Calculus
Recent Progress on a Statistical Network Calculus Jorg Liebeherr Deartment of Comuter Science University of Virginia Collaborators Almut Burchard Robert Boorstyn Chaiwat Oottamakorn Stehen Patek Chengzhi
More informationCPU SCHEDULING RONG ZHENG
CPU SCHEDULING RONG ZHENG OVERVIEW Why scheduling? Non-preemptive vs Preemptive policies FCFS, SJF, Round robin, multilevel queues with feedback, guaranteed scheduling 2 SHORT-TERM, MID-TERM, LONG- TERM
More informationChapter 2 Queueing Theory and Simulation
Chapter 2 Queueing Theory and Simulation Based on the slides of Dr. Dharma P. Agrawal, University of Cincinnati and Dr. Hiroyuki Ohsaki Graduate School of Information Science & Technology, Osaka University,
More informationAS computer hardware technology advances, both
1 Best-Harmonically-Fit Periodic Task Assignment Algorithm on Multiple Periodic Resources Chunhui Guo, Student Member, IEEE, Xiayu Hua, Student Member, IEEE, Hao Wu, Student Member, IEEE, Douglas Lautner,
More informationOn a Closed Formula for the Derivatives of e f(x) and Related Financial Applications
International Mathematical Forum, 4, 9, no. 9, 41-47 On a Closed Formula or the Derivatives o e x) and Related Financial Applications Konstantinos Draais 1 UCD CASL, University College Dublin, Ireland
More informationA POPULATION-MIX DRIVEN APPROXIMATION FOR QUEUEING NETWORKS WITH FINITE CAPACITY REGIONS
A POPULATION-MIX DRIVEN APPROXIMATION FOR QUEUEING NETWORKS WITH FINITE CAPACITY REGIONS J. Anselmi 1, G. Casale 2, P. Cremonesi 1 1 Politecnico di Milano, Via Ponzio 34/5, I-20133 Milan, Italy 2 Neptuny
More informationEmbedded Systems 14. Overview of embedded systems design
Embedded Systems 14-1 - Overview of embedded systems design - 2-1 Point of departure: Scheduling general IT systems In general IT systems, not much is known about the computational processes a priori The
More informationAn Improved Bound for Minimizing the Total Weighted Completion Time of Coflows in Datacenters
IEEE/ACM TRANSACTIONS ON NETWORKING An Improved Bound for Minimizing the Total Weighted Completion Time of Coflows in Datacenters Mehrnoosh Shafiee, Student Member, IEEE, and Javad Ghaderi, Member, IEEE
More informationNATCOR: Stochastic Modelling
NATCOR: Stochastic Modelling Queueing Theory II Chris Kirkbride Management Science 2017 Overview of Today s Sessions I Introduction to Queueing Modelling II Multiclass Queueing Models III Queueing Control
More informationCPU scheduling. CPU Scheduling
EECS 3221 Operating System Fundamentals No.4 CPU scheduling Prof. Hui Jiang Dept of Electrical Engineering and Computer Science, York University CPU Scheduling CPU scheduling is the basis of multiprogramming
More information1 Introduction Future high speed digital networks aim to serve integrated trac, such as voice, video, fax, and so forth. To control interaction among
On Deterministic Trac Regulation and Service Guarantees: A Systematic Approach by Filtering Cheng-Shang Chang Dept. of Electrical Engineering National Tsing Hua University Hsinchu 30043 Taiwan, R.O.C.
More informationWhen Periodic Timetables are Suboptimal
Konrad-Zuse-Zentrum ür Inormationstechnik Berlin akustraße 7 D-14195 Berlin-Dahlem Germany RALF BORNDÖRFER CHRISIAN LIEBCHEN When Periodic imetables are Suboptimal his work was unded by the DFG Research
More informationPerformance Evaluation of Queuing Systems
Performance Evaluation of Queuing Systems Introduction to Queuing Systems System Performance Measures & Little s Law Equilibrium Solution of Birth-Death Processes Analysis of Single-Station Queuing Systems
More informationCSCE 313 Introduction to Computer Systems. Instructor: Dezhen Song
CSCE 313 Introduction to Computer Systems Instructor: Dezhen Song Schedulers in the OS CPU Scheduling Structure of a CPU Scheduler Scheduling = Selection + Dispatching Criteria for scheduling Scheduling
More informationTESTING TIMED FINITE STATE MACHINES WITH GUARANTEED FAULT COVERAGE
TESTING TIMED FINITE STATE MACHINES WITH GUARANTEED FAULT COVERAGE Khaled El-Fakih 1, Nina Yevtushenko 2 *, Hacene Fouchal 3 1 American University o Sharjah, PO Box 26666, UAE kelakih@aus.edu 2 Tomsk State
More informationTDDI04, K. Arvidsson, IDA, Linköpings universitet CPU Scheduling. Overview: CPU Scheduling. [SGG7] Chapter 5. Basic Concepts.
TDDI4 Concurrent Programming, Operating Systems, and Real-time Operating Systems CPU Scheduling Overview: CPU Scheduling CPU bursts and I/O bursts Scheduling Criteria Scheduling Algorithms Multiprocessor
More informationAndrew Morton University of Waterloo Canada
EDF Feasibility and Hardware Accelerators Andrew Morton University of Waterloo Canada Outline 1) Introduction and motivation 2) Review of EDF and feasibility analysis 3) Hardware accelerators and scheduling
More informationSupplementary material for Continuous-action planning for discounted infinite-horizon nonlinear optimal control with Lipschitz values
Supplementary material or Continuous-action planning or discounted ininite-horizon nonlinear optimal control with Lipschitz values List o main notations x, X, u, U state, state space, action, action space,
More informationTelescoping Decomposition Method for Solving First Order Nonlinear Differential Equations
Telescoping Decomposition Method or Solving First Order Nonlinear Dierential Equations 1 Mohammed Al-Reai 2 Maysem Abu-Dalu 3 Ahmed Al-Rawashdeh Abstract The Telescoping Decomposition Method TDM is a new
More informationSchedulability of Periodic and Sporadic Task Sets on Uniprocessor Systems
Schedulability of Periodic and Sporadic Task Sets on Uniprocessor Systems Jan Reineke Saarland University July 4, 2013 With thanks to Jian-Jia Chen! Jan Reineke July 4, 2013 1 / 58 Task Models and Scheduling
More informationTardiness Bounds under Global EDF Scheduling on a. Multiprocessor
Tardiness Bounds under Global EDF Scheduling on a Multiprocessor UmaMaheswari C. Devi and James H. Anderson Department of Computer Science The University of North Carolina at Chapel Hill Abstract We consider
More informationCPSC 531: System Modeling and Simulation. Carey Williamson Department of Computer Science University of Calgary Fall 2017
CPSC 531: System Modeling and Simulation Carey Williamson Department of Computer Science University of Calgary Fall 2017 Motivating Quote for Queueing Models Good things come to those who wait - poet/writer
More informationThe Rate-Based Execution Model
University of Nebraska - Lincoln DigitalCommons@University of Nebraska - Lincoln CSE Technical reports Computer Science and Engineering, Department of 4-1-1999 The Rate-Based Execution Model Kevin Jeffay
More informationOn the Flow-level Dynamics of a Packet-switched Network
On the Flow-level Dynamics o a Packet-switched Network Ciamac C. Moallemi and Devavrat Shah Abstract: The packet is the undamental unit o transportation in modern communication networks such as the Internet.
More informationScheduling. Uwe R. Zimmer & Alistair Rendell The Australian National University
6 Scheduling Uwe R. Zimmer & Alistair Rendell The Australian National University References for this chapter [Bacon98] J. Bacon Concurrent Systems 1998 (2nd Edition) Addison Wesley Longman Ltd, ISBN 0-201-17767-6
More informationCapacity and Scheduling in Small-Cell HetNets
Capacity and Scheduling in Small-Cell HetNets Stephen Hanly Macquarie University North Ryde, NSW 2109 Joint Work with Sem Borst, Chunshan Liu and Phil Whiting Tuesday, January 13, 2015 Hanly Capacity and
More informationQuality of Real-Time Streaming in Wireless Cellular Networks : Stochastic Modeling and Analysis
Quality of Real-Time Streaming in Wireless Cellular Networs : Stochastic Modeling and Analysis B. Blaszczyszyn, M. Jovanovic and M. K. Karray Based on paper [1] WiOpt/WiVid Mai 16th, 2014 Outline Introduction
More informationDelay Efficient Scheduling via Redundant Constraints in Multihop Networks
Delay Eicient Scheduling via Redundant Constraints in Multihop Networks Longbo Huang, Michael J. Neely Abstract We consider the problem o delay-eicient scheduling in general multihop networks. While the
More informationThe Concurrent Consideration of Uncertainty in WCETs and Processor Speeds in Mixed Criticality Systems
The Concurrent Consideration of Uncertainty in WCETs and Processor Speeds in Mixed Criticality Systems Zhishan Guo and Sanjoy Baruah Department of Computer Science University of North Carolina at Chapel
More informationNumerical Methods - Lecture 2. Numerical Methods. Lecture 2. Analysis of errors in numerical methods
Numerical Methods - Lecture 1 Numerical Methods Lecture. Analysis o errors in numerical methods Numerical Methods - Lecture Why represent numbers in loating point ormat? Eample 1. How a number 56.78 can
More informationNetworked Embedded Systems WS 2016/17
Networked Embedded Systems WS 2016/17 Lecture 2: Real-time Scheduling Marco Zimmerling Goal of Today s Lecture Introduction to scheduling of compute tasks on a single processor Tasks need to finish before
More informationQueueing Systems: Lecture 3. Amedeo R. Odoni October 18, Announcements
Queueing Systems: Lecture 3 Amedeo R. Odoni October 18, 006 Announcements PS #3 due tomorrow by 3 PM Office hours Odoni: Wed, 10/18, :30-4:30; next week: Tue, 10/4 Quiz #1: October 5, open book, in class;
More informationCPU Scheduling. CPU Scheduler
CPU Scheduling These slides are created by Dr. Huang of George Mason University. Students registered in Dr. Huang s courses at GMU can make a single machine readable copy and print a single copy of each
More informationSection 1.2: A Single Server Queue
Section 12: A Single Server Queue Discrete-Event Simulation: A First Course c 2006 Pearson Ed, Inc 0-13-142917-5 Discrete-Event Simulation: A First Course Section 12: A Single Server Queue 1/ 30 Section
More informationFair Operation of Multi-Server and Multi-Queue Systems
Fair Operation of Multi-Server and Multi-Queue Systems David Raz School of Computer Science Tel-Aviv University, Tel-Aviv, Israel davidraz@post.tau.ac.il Benjamin Avi-Itzhak RUTCOR, Rutgers University,
More informationControl of Fork-Join Networks in Heavy-Traffic
in Heavy-Traffic Asaf Zviran Based on MSc work under the guidance of Rami Atar (Technion) and Avishai Mandelbaum (Technion) Industrial Engineering and Management Technion June 2010 Introduction Network
More informationDeadline-driven scheduling
Deadline-driven scheduling Michal Sojka Czech Technical University in Prague, Faculty of Electrical Engineering, Department of Control Engineering November 8, 2017 Some slides are derived from lectures
More informationAnalyzing Queuing Systems with Coupled Processors. through Semidefinite Programming
Analyzing Queuing Systems with Coupled Processors through Semidefinite Programming Balaji Rengarajan, Constantine Caramanis, and Gustavo de Veciana Dept. of Electrical and Computer Engineering The University
More informationPhysics 5153 Classical Mechanics. Solution by Quadrature-1
October 14, 003 11:47:49 1 Introduction Physics 5153 Classical Mechanics Solution by Quadrature In the previous lectures, we have reduced the number o eective degrees o reedom that are needed to solve
More informationQueueing Theory I Summary! Little s Law! Queueing System Notation! Stationary Analysis of Elementary Queueing Systems " M/M/1 " M/M/m " M/M/1/K "
Queueing Theory I Summary Little s Law Queueing System Notation Stationary Analysis of Elementary Queueing Systems " M/M/1 " M/M/m " M/M/1/K " Little s Law a(t): the process that counts the number of arrivals
More informationSchedulability analysis of global Deadline-Monotonic scheduling
Schedulability analysis of global Deadline-Monotonic scheduling Sanjoy Baruah Abstract The multiprocessor Deadline-Monotonic (DM) scheduling of sporadic task systems is studied. A new sufficient schedulability
More informationResource Allocation for Video Streaming in Wireless Environment
Resource Allocation for Video Streaming in Wireless Environment Shahrokh Valaee and Jean-Charles Gregoire Abstract This paper focuses on the development of a new resource allocation scheme for video streaming
More informationCSE 380 Computer Operating Systems
CSE 380 Computer Operating Systems Instructor: Insup Lee & Dianna Xu University of Pennsylvania, Fall 2003 Lecture Note 3: CPU Scheduling 1 CPU SCHEDULING q How can OS schedule the allocation of CPU cycles
More informationLecture 13. Real-Time Scheduling. Daniel Kästner AbsInt GmbH 2013
Lecture 3 Real-Time Scheduling Daniel Kästner AbsInt GmbH 203 Model-based Software Development 2 SCADE Suite Application Model in SCADE (data flow + SSM) System Model (tasks, interrupts, buses, ) SymTA/S
More informationMODELLING PUBLIC TRANSPORT CORRIDORS WITH AGGREGATE AND DISAGGREGATE DEMAND
MODELLIG PUBLIC TRASPORT CORRIDORS WITH AGGREGATE AD DISAGGREGATE DEMAD Sergio Jara-Díaz Alejandro Tirachini and Cristián Cortés Universidad de Chile ITRODUCTIO Traditional microeconomic modeling o public
More informationGiuseppe Bianchi, Ilenia Tinnirello
Capacity of WLAN Networs Summary Per-node throughput in case of: Full connected networs each node sees all the others Generic networ topology not all nodes are visible Performance Analysis of single-hop
More informationCPU Scheduling Exercises
CPU Scheduling Exercises NOTE: All time in these exercises are in msec. Processes P 1, P 2, P 3 arrive at the same time, but enter the job queue in the order presented in the table. Time quantum = 3 msec
More informationEmbedded Systems Development
Embedded Systems Development Lecture 3 Real-Time Scheduling Dr. Daniel Kästner AbsInt Angewandte Informatik GmbH kaestner@absint.com Model-based Software Development Generator Lustre programs Esterel programs
More informationOn High-Rate Cryptographic Compression Functions
On High-Rate Cryptographic Compression Functions Richard Ostertág and Martin Stanek Department o Computer Science Faculty o Mathematics, Physics and Inormatics Comenius University Mlynská dolina, 842 48
More informationSimulation of Process Scheduling Algorithms
Simulation of Process Scheduling Algorithms Project Report Instructor: Dr. Raimund Ege Submitted by: Sonal Sood Pramod Barthwal Index 1. Introduction 2. Proposal 3. Background 3.1 What is a Process 4.
More informationMinimizing the Number of Tardy Jobs
Minimizing the Number of Tardy Jobs 1 U j Example j p j d j 1 10 10 2 2 11 3 7 13 4 4 15 5 8 20 Ideas: Need to choose a subset of jobs S that meet their deadlines. Schedule the jobs that meet their deadlines
More informationIntroduction to queuing theory
Introduction to queuing theory Claude Rigault ENST claude.rigault@enst.fr Introduction to Queuing theory 1 Outline The problem The number of clients in a system The client process Delay processes Loss
More informationA Mechanism for Pricing Service Guarantees
A Mechanism for Pricing Service Guarantees Bruce Hajek Department of Electrical and Computer Engineering and the Coordinated Science Laboratory University of Illinois at Urbana-Champaign Sichao Yang Qualcomm
More informationDynamic resource sharing
J. Virtamo 38.34 Teletraffic Theory / Dynamic resource sharing and balanced fairness Dynamic resource sharing In previous lectures we have studied different notions of fair resource sharing. Our focus
More informationLIMITS FOR QUEUES AS THE WAITING ROOM GROWS. Bell Communications Research AT&T Bell Laboratories Red Bank, NJ Murray Hill, NJ 07974
LIMITS FOR QUEUES AS THE WAITING ROOM GROWS by Daniel P. Heyman Ward Whitt Bell Communications Research AT&T Bell Laboratories Red Bank, NJ 07701 Murray Hill, NJ 07974 May 11, 1988 ABSTRACT We study the
More informationCompetitive Management of Non-Preemptive Queues with Multiple Values
Competitive Management of Non-Preemptive Queues with Multiple Values Nir Andelman and Yishay Mansour School of Computer Science, Tel-Aviv University, Tel-Aviv, Israel Abstract. We consider the online problem
More information