Jackson network

Jackson network (Redirected from Jackson's theorem (queueing theory)) Jump to navigation Jump to search This article includes a list of general references, aber es fehlen genügend entsprechende Inline-Zitate. Bitte helfen Sie mit, diesen Artikel zu verbessern, indem Sie genauere Zitate einfügen. (Juni 2012) (Erfahren Sie, wie und wann Sie diese Vorlagennachricht entfernen können) In queueing theory, a discipline within the mathematical theory of probability, a Jackson network (sometimes Jacksonian network[1]) is a class of queueing network where the equilibrium distribution is particularly simple to compute as the network has a product-form solution. It was the first significant development in the theory of networks of queues, and generalising and applying the ideas of the theorem to search for similar product-form solutions in other networks has been the subject of much research,[2] including ideas used in the development of the Internet.[3] The networks were first identified by James R. Jackson[4][5] and his paper was re-printed in the journal Management Science’s ‘Ten Most Influential Titles of Management Sciences First Fifty Years.’[6] Jackson was inspired by the work of Burke and Reich,[7] though Jean Walrand notes "product-form results … [sind] a much less immediate result of the output theorem than Jackson himself appeared to believe in his fundamental paper".[8] An earlier product-form solution was found by R. R. P. Jackson for tandem queues (a finite chain of queues where each customer must visit each queue in order) and cyclic networks (a loop of queues where each customer must visit each queue in order).[9] A Jackson network consists of a number of nodes, where each node represents a queue in which the service rate can be both node-dependent (different nodes have different service rates) and state-dependent (service rates change depending on queue lengths). Jobs travel among the nodes following a fixed routing matrix. All jobs at each node belong to a single "Klasse" and jobs follow the same service-time distribution and the same routing mechanism. Folglich, there is no notion of priority in serving the jobs: all jobs at each node are served on a first-come, first-served basis.

Jackson networks where a finite population of jobs travel around a closed network also have a product-form solution described by the Gordon–Newell theorem.[10] Inhalt 1 Necessary conditions for a Jackson network 2 Satz 3 Definition 3.1 Satz 3.2 Beispiel 4 Generalized Jackson network 4.1 Brownian approximation 5 Siehe auch 6 References Necessary conditions for a Jackson network A network of m interconnected queues is known as a Jackson network[11] or Jacksonian network[12] if it meets the following conditions: if the network is open, any external arrivals to node i form a Poisson process, All service times are exponentially distributed and the service discipline at all queues is first-come, first-served, a customer completing service at queue i will either move to some new queue j with probability {Anzeigestil P_{ij}} or leave the system with probability {displaystyle 1-sum _{j=1}^{m}P_{ij}} , die, for an open network, is non-zero for some subset of the queues, the utilization of all of the queues is less than one. Theorem In an open Jackson network of m M/M/1 queues where the utilization {Anzeigestil rho _{ich}} is less than 1 at every queue, the equilibrium state probability distribution exists and for state {displaystyle scriptstyle {(k_{1},k_{2},Punkte ,k_{m})}} is given by the product of the individual queue equilibrium distributions {Anzeigestil pi (k_{1},k_{2},Punkte ,k_{m})=prod _{i=1}^{m}Pi _{ich}(k_{ich})=prod _{i=1}^{m}[rho_{ich}^{k_{ich}}(1-rho_{ich})].} The result {Anzeigestil pi (k_{1},k_{2},Punkte ,k_{m})=prod _{i=1}^{m}Pi _{ich}(k_{ich})} also holds for M/M/c model stations with ci servers at the {displaystyle i^{Text{th}}} station, with utilization requirement {Anzeigestil rho _{ich}0} . Each arrival is independently routed to node j with probability {Anzeigestil p_{0j}geq 0} und {Anzeigestil Summe _{j=1}^{J}p_{0j}=1} . Upon service completion at node i, a job may go to another node j with probability {Anzeigestil p_{ij}} or leave the network with probability {Anzeigestil p_{i0}=1-Summe _{j=1}^{J}p_{ij}} .

Hence we have the overall arrival rate to node i, {Anzeigestil Lambda _{ich}} , including both external arrivals and internal transitions: {Anzeigestil Lambda _{ich}=alpha p_{0ich}+Summe _{j=1}^{J}Lambda _{j}p_{ji},i=1,lPunkte ,J.qquad (1)} (Since the utilisation at each node is less than 1, and we are looking at the equilibrium distribution i.e. the long-run-average behaviour, the rate of jobs transitioning from j to i is bounded by a fraction of the arrival rate at j and we ignore the service rate {displaystyle ihn _{j}} in the above.) Definieren {displaystyle a=(alpha p_{0ich})_{i=1}^{J}} , then we can solve {Anzeigestil Lambda =(I-P^{T})^{-1}a} .

All jobs leave each node also following Poisson process, und definieren {displaystyle ihn _{ich}(x_{ich})} as the service rate of node i when there are {Anzeigestil x_{ich}} jobs at node i.

Lassen {Anzeigestil X_{ich}(t)} denote the number of jobs at node i at time t, und {Anzeigestil mathbf {X} =(X_{ich})_{i=1}^{J}} . Then the equilibrium distribution of {Anzeigestil mathbf {X} } , {Anzeigestil pi (mathbf {x} )=P(mathbf {X} =mathbf {x} )} is determined by the following system of balance equations: {Anzeigestil {Start{ausgerichtet}&pi (mathbf {x} )Summe _{i=1}^{J}[alpha p_{0ich}+in _{ich}(x_{ich})(1-p_{ii})]\={}&sum _{i=1}^{J}[Pi (mathbf {x} -mathbf {e} _{ich})alpha p_{0ich}+Pi (mathbf {x} +mathbf {e} _{ich})in _{ich}(x_{ich}+1)p_{i0}]+Summe _{i=1}^{J}Summe _{jneq i}Pi (mathbf {x} +mathbf {e} _{ich}-mathbf {e} _{j})in _{ich}(x_{ich}+1)p_{ij}.Quad (2)Ende{ausgerichtet}}} wo {Anzeigestil mathbf {e} _{ich}} denote the {displaystyle i^{Text{th}}} unit vector.

Theorem Suppose a vector of independent random variables {Anzeigestil (Y_{1},Punkte ,Y_{J})} with each {Anzeigestil Y_{ich}} having a probability mass function as {Anzeigestil P(Y_{ich}=n)=p(Y_{ich}=0)cdot {frac {Lambda _{ich}^{n}}{M_{ich}(n)}},Quad (3)} wo {Anzeigestil M_{ich}(n)=prod _{j=1}^{n}in _{ich}(j)} . Wenn {Anzeigestil Summe _{n=1}^{unendlich }{frac {Lambda _{ich}^{n}}{M_{ich}(n)}}0} Then by the theorem, we can calculate: {Anzeigestil Lambda =(I-P^{T})^{-1}a={Start{bMatrix}1&0&0\-0.5&1&0\-0.5&0&1end{bMatrix}}^{-1}{Start{bMatrix}0.5times 5\0.5times 5\0end{bMatrix}}={Start{bMatrix}1&0&0\0.5&1&0\0.5&0&1end{bMatrix}}{Start{bMatrix}2.5\2.5\0Ende{bMatrix}}={Start{bMatrix}2.5\3.75\1.25Ende{bMatrix}}} According to the definition of {Anzeigestil mathbf {Y} } , wir haben: {Anzeigestil P(Y_{1}=0)=links(Summe _{n=0}^{unendlich }links({frac {2.5}{15}}Rechts)^{n}Rechts)^{-1}={frac {5}{6}}} {Anzeigestil P(Y_{2}=0)=links(Summe _{n=0}^{unendlich }links({frac {3.75}{12}}Rechts)^{n}Rechts)^{-1}={frac {11}{16}}} {Anzeigestil P(Y_{3}=0)=links(Summe _{n=0}^{unendlich }links({frac {1.25}{10}}Rechts)^{n}Rechts)^{-1}={frac {7}{8}}} Hence the probability that there is one job at each node is: {Anzeigestil pi (1,1,1)={frac {5}{6}}cdot {frac {2.5}{15}}cdot {frac {11}{16}}cdot {frac {3.75}{12}}cdot {frac {7}{8}}cdot {frac {1.25}{10}}ca 0.00326} Since the service rate here does not depend on state, das {Anzeigestil Y_{ich}} s simply follow a geometric distribution.

Generalized Jackson network A generalized Jackson network allows renewal arrival processes that need not be Poisson processes, and independent, identically distributed non-exponential service times. Im Algemeinen, this network does not have a product-form stationary distribution, so approximations are sought.[13] Brownian approximation Under some mild conditions the queue-length process[Klärung nötig] {Anzeigestil Q(t)} of an open generalized Jackson network can be approximated by a reflected Brownian motion defined as {Anzeigestil Betreibername {RBM} _{Q(0)}(Theta ,Gamma ;R).} , wo {Theta im Display-Stil } is the drift of the process, {Anzeigestil Gamma } is the covariance matrix, und {Anzeigestil R} is the reflection matrix. This is a two-order approximation obtained by relation between general Jackson network with homogeneous fluid network and reflected Brownian motion.

The parameters of the reflected Brownian process is specified as follows: {displaystyle theta =alpha -(I-P^{T})in } {displaystyle Gamma =(Gamma _{kell }){Text{ mit }}Gamma _{kell }= Summe _{j=1}^{J}(Lambda _{j}wedge mu _{j})[p_{jk}(Delta _{kell }-p_{jell })+c_{j}^{2}(p_{jk}-Delta _{jk})(p_{jell }-Delta _{jell })]+Alpha _{k}c_{0,k}^{2}Delta _{kell }} {displaystyle R=I-P^{T}} where the symbols are defined as: Definitions of symbols in the approximation formula symbol Meaning {displaystyle alpha =(Alpha _{j})_{j=1}^{J}} a J-vector specifying the arrival rates to each node.

{displaystyle mu =(in )_{j=1}^{J}} a J-vector specifying the service rates of each node.

{Anzeigestil P} routing matrix.

{Anzeigestil Lambda _{j}} effective arrival of {displaystyle j^{Text{th}}} node.

{Anzeigestil c_{j}} variation of service time at {displaystyle j^{Text{th}}} node.

{Anzeigestil c_{0,j}} variation of inter-arrival time at {displaystyle j^{Text{th}}} node.

{displaystyle delta _{ij}} coefficients to specify correlation between nodes. show See also Gordon–Newell network BCMP network G-network Little's law References ^ Walrand, J.; Varaiya, P. (1980). "Sojourn Times and the Overtaking Condition in Jacksonian Networks". Advances in Applied Probability. 12 (4): 1000–1018. doi:10.2307/1426753. JSTOR 1426753. ^ Kelly, F. P. (Juni 1976). "Networks of Queues". Advances in Applied Probability. 8 (2): 416–432. doi:10.2307/1425912. JSTOR 1425912. ^ Jackson, Jakob R. (Dezember 2004). "Comments on "Jobshop-Like Queueing Systems": The Background". Management Science. 50 (12): 1796–1802. doi:10.1287/mnsc.1040.0268. JSTOR 30046150. ^ Jackson, Jakob R. (Oct 1963). "Jobshop-like Queueing Systems". Management Science. 10 (1): 131–142. doi:10.1287/mnsc.1040.0268. JSTOR 2627213. A version from January 1963 is available at http://www.dtic.mil/dtic/tr/fulltext/u2/296776.pdf ^ Jackson, J. R. (1957). "Networks of Waiting Lines". Operations Research. 5 (4): 518–521. doi:10.1287/opre.5.4.518. JSTOR 167249. ^ Jackson, Jakob R. (Dezember 2004). "Jobshop-Like Queueing Systems". Management Science. 50 (12): 1796–1802. doi:10.1287/mnsc.1040.0268. JSTOR 30046149. ^ Reich, Edgar (September 1957). "Waiting Times When Queues are in Tandem". Annals of Mathematical Statistics. 28 (3): 768. doi:10.1214/aoms/1177706889. JSTOR 2237237. ^ Walrand, Jean (November 1983). "A Probabilistic Look at Networks of Quasi-Reversible Queues". IEEE Transactions on Information Theory. 29 (6): 825. doi:10.1109/TIT.1983.1056762. ^ Jackson, R. R. P. (1995). "Book review: Queueing networks and product forms: a systems approach". IMA Journal of Management Mathematics. 6 (4): 382–384. doi:10.1093/imaman/6.4.382. ^ Gordon, W. J.; Newell, G. F. (1967). "Closed Queuing Systems with Exponential Servers". Operations Research. 15 (2): 254. doi:10.1287/opre.15.2.254. JSTOR 168557. ^ Goodman, Jonathan B.; Massey, William A. (Dezember 1984). "The Non-Ergodic Jackson Network". Journal of Applied Probability. 21 (4): 860–869. doi:10.2307/3213702. ^ Walrand, J.; Varaiya, P. (Dezember 1980). "Sojourn Times and the Overtaking Condition in Jacksonian Networks". Advances in Applied Probability. 12 (4): 1000–1018. doi:10.2307/1426753. ^ Chen, Hong; Yao, David D. (2001). Fundamentals of Queueing Networks: Performance, Asymptotics, and Optimization. Springer. ISBN 0-387-95166-0. hide vte Queueing theory Single queueing nodes D/M/1 queueM/D/1 queueM/D/c queueM/M/1 queue Burke's theoremM/M/c queueM/M/∞ queueM/G/1 queue Pollaczek–Khinchine formulaMatrix analytic methodM/G/k queueG/M/1 queueG/G/1 queue Kingman's formulaLindley equationFork–join queueBulk queue Arrival processes Poisson point processMarkovian arrival processRational arrival process Queueing networks Jackson network Traffic equationsGordon–Newell theorem Mean value analysisBuzen's algorithmKelly networkG-networkBCMP network Service policies FIFOLIFOProcessor sharingRound-robinShortest job nextShortest remaining time Key concepts Continuous-time Markov chainKendall's notationLittle's lawProduct-form solution Balance equationQuasireversibilityFlow-equivalent server methodArrival theoremDecomposition methodBeneš method Limit theorems Fluid limitMean-field theoryHeavy traffic approximation Reflected Brownian motion Extensions Fluid queueLayered queueing networkPolling systemAdversarial queueing networkLoss networkRetrial queue Information systems Data bufferErlang (Einheit)Erlang distributionFlow control (data)Message queueNetwork congestionNetwork schedulerPipeline (software)Quality of serviceScheduling (computing)Teletraffic engineering Category Categories: Queueing theory