Approximation of Excessive Backlog Probabilities of Two Tandem Queues

Let $X$ be the constrained random walk on ${\mathbb Z}_+^2$ taking the steps $(1,0)$, $(-1,1)$ and $(0,-1)$ with probabilities $\lambda < (\mu_1\neq \mu_2)$; in particular, $X$ is assumed stable. Let $\tau_n$ be the first time $X$ hits $\partial A_n = \{x:x(1)+x(2) = n \}$ For $x \in {\mathbb Z}_...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
1. Verfasser: Sezer, Ali Devin
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page
container_issue
container_start_page
container_title
container_volume
creator Sezer, Ali Devin
description Let $X$ be the constrained random walk on ${\mathbb Z}_+^2$ taking the steps $(1,0)$, $(-1,1)$ and $(0,-1)$ with probabilities $\lambda < (\mu_1\neq \mu_2)$; in particular, $X$ is assumed stable. Let $\tau_n$ be the first time $X$ hits $\partial A_n = \{x:x(1)+x(2) = n \}$ For $x \in {\mathbb Z}_+^2, x(1) + x(2) < n$, the probability $p_n(x)= P_x( \tau_n < \tau_0)$ is a key performance measure for the queueing system represented by $X$. Let $Y$ be the constrained random walk on ${\mathbb Z} \times {\mathbb Z}_+$ with increments $(-1,0)$, $(1,1)$ and $(0,-1)$. Let $\tau$ be the first time that the components of $Y$ equal each other. We derive the following explicit formula for $P_y(\tau < \infty)$: \[ P_y(\tau < \infty) = W(y)= \rho_2^{y(1)-y(2)} + \frac{\mu_2 - \lambda}{\mu_2 - \mu_1} \rho_1^{ y(1)-y(2)} \rho_1^{y(2)} + \frac{\mu_2-\lambda}{\mu_1 -\mu_2} \rho_2^{y(1)-y(2)} \rho_1^{y(2)}, \] where, $\rho_i = \lambda/\mu_i$, $i=1,2$, $y \in {\mathbb Z}\times{ \mathbb Z}_+$, $y(1) > y(2)$, and show that $W(n-x_n(1),x_n(2))$ approximates $p_n(x_n)$ with relative error {\em exponentially decaying} in $n$ for $x_n = \lfloor nx \rfloor$, $x \in {\mathbb R}_+^2$, $0 < x(1) + x(2) < 1$. The steps of our analysis: 1) with an affine transformation, move the origin $(0,0)$ to $(n,0)$ on $\partial A_n$; let $n\nearrow \infty$ to remove the constraint on the $x(2)$ axis; this step gives the limit {\em unstable} /{\em transient} constrained random walk $Y$ and reduces $P_{x}(\tau_n < \tau_0)$ to $P_y(\tau < \infty)$; 2) construct a basis of harmonic functions of $Y$ and use it to apply the superposition principle to compute $P_y(\tau < \infty).$ The construction involves the use of conjugate points on a characteristic surface associated with the walk $X$. The proof that the relative error decays exponentially uses a sequence of subsolutions of a related HJB equation on a manifold.
doi_str_mv 10.48550/arxiv.1801.04674
format Article
fullrecord <record><control><sourceid>arxiv_GOX</sourceid><recordid>TN_cdi_arxiv_primary_1801_04674</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>1801_04674</sourcerecordid><originalsourceid>FETCH-LOGICAL-a674-894fbc77d8349ffade4f333d5123ff42cefcba9dd5063aeb47765225ad7ceb173</originalsourceid><addsrcrecordid>eNotj8tOwzAURL1hURU-gBX-gQQ7tuNkWarwkCoBUvbRtX2NLNI6stsS_p62dDWb0Zw5hNxzVspGKfYIaQ7HkjeMl0zWWi5It5qmFOewhX2IOxo97WaLOYcj0iew32P8oh8pGjBhDPuA-VzpfyLtYedwSz8PeMB8S248jBnvrrkk_XPXr1-LzfvL23q1KeAEK5pWemO1do2QrffgUHohhFO8Et7LyqK3BlrnFKsFoJFa16qqFDht0XAtluThf_biMUzpdDv9Dmef4eIj_gABZ0aW</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>Approximation of Excessive Backlog Probabilities of Two Tandem Queues</title><source>arXiv.org</source><creator>Sezer, Ali Devin</creator><creatorcontrib>Sezer, Ali Devin</creatorcontrib><description><![CDATA[Let $X$ be the constrained random walk on ${\mathbb Z}_+^2$ taking the steps $(1,0)$, $(-1,1)$ and $(0,-1)$ with probabilities $\lambda < (\mu_1\neq \mu_2)$; in particular, $X$ is assumed stable. Let $\tau_n$ be the first time $X$ hits $\partial A_n = \{x:x(1)+x(2) = n \}$ For $x \in {\mathbb Z}_+^2, x(1) + x(2) < n$, the probability $p_n(x)= P_x( \tau_n < \tau_0)$ is a key performance measure for the queueing system represented by $X$. Let $Y$ be the constrained random walk on ${\mathbb Z} \times {\mathbb Z}_+$ with increments $(-1,0)$, $(1,1)$ and $(0,-1)$. Let $\tau$ be the first time that the components of $Y$ equal each other. We derive the following explicit formula for $P_y(\tau < \infty)$: \[ P_y(\tau < \infty) = W(y)= \rho_2^{y(1)-y(2)} + \frac{\mu_2 - \lambda}{\mu_2 - \mu_1} \rho_1^{ y(1)-y(2)} \rho_1^{y(2)} + \frac{\mu_2-\lambda}{\mu_1 -\mu_2} \rho_2^{y(1)-y(2)} \rho_1^{y(2)}, \] where, $\rho_i = \lambda/\mu_i$, $i=1,2$, $y \in {\mathbb Z}\times{ \mathbb Z}_+$, $y(1) > y(2)$, and show that $W(n-x_n(1),x_n(2))$ approximates $p_n(x_n)$ with relative error {\em exponentially decaying} in $n$ for $x_n = \lfloor nx \rfloor$, $x \in {\mathbb R}_+^2$, $0 < x(1) + x(2) < 1$. The steps of our analysis: 1) with an affine transformation, move the origin $(0,0)$ to $(n,0)$ on $\partial A_n$; let $n\nearrow \infty$ to remove the constraint on the $x(2)$ axis; this step gives the limit {\em unstable} /{\em transient} constrained random walk $Y$ and reduces $P_{x}(\tau_n < \tau_0)$ to $P_y(\tau < \infty)$; 2) construct a basis of harmonic functions of $Y$ and use it to apply the superposition principle to compute $P_y(\tau < \infty).$ The construction involves the use of conjugate points on a characteristic surface associated with the walk $X$. The proof that the relative error decays exponentially uses a sequence of subsolutions of a related HJB equation on a manifold.]]></description><identifier>DOI: 10.48550/arxiv.1801.04674</identifier><language>eng</language><subject>Mathematics - Probability</subject><creationdate>2018-01</creationdate><rights>http://arxiv.org/licenses/nonexclusive-distrib/1.0</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>228,230,781,886</link.rule.ids><linktorsrc>$$Uhttps://arxiv.org/abs/1801.04674$$EView_record_in_Cornell_University$$FView_record_in_$$GCornell_University$$Hfree_for_read</linktorsrc><backlink>$$Uhttps://doi.org/10.48550/arXiv.1801.04674$$DView paper in arXiv$$Hfree_for_read</backlink></links><search><creatorcontrib>Sezer, Ali Devin</creatorcontrib><title>Approximation of Excessive Backlog Probabilities of Two Tandem Queues</title><description><![CDATA[Let $X$ be the constrained random walk on ${\mathbb Z}_+^2$ taking the steps $(1,0)$, $(-1,1)$ and $(0,-1)$ with probabilities $\lambda < (\mu_1\neq \mu_2)$; in particular, $X$ is assumed stable. Let $\tau_n$ be the first time $X$ hits $\partial A_n = \{x:x(1)+x(2) = n \}$ For $x \in {\mathbb Z}_+^2, x(1) + x(2) < n$, the probability $p_n(x)= P_x( \tau_n < \tau_0)$ is a key performance measure for the queueing system represented by $X$. Let $Y$ be the constrained random walk on ${\mathbb Z} \times {\mathbb Z}_+$ with increments $(-1,0)$, $(1,1)$ and $(0,-1)$. Let $\tau$ be the first time that the components of $Y$ equal each other. We derive the following explicit formula for $P_y(\tau < \infty)$: \[ P_y(\tau < \infty) = W(y)= \rho_2^{y(1)-y(2)} + \frac{\mu_2 - \lambda}{\mu_2 - \mu_1} \rho_1^{ y(1)-y(2)} \rho_1^{y(2)} + \frac{\mu_2-\lambda}{\mu_1 -\mu_2} \rho_2^{y(1)-y(2)} \rho_1^{y(2)}, \] where, $\rho_i = \lambda/\mu_i$, $i=1,2$, $y \in {\mathbb Z}\times{ \mathbb Z}_+$, $y(1) > y(2)$, and show that $W(n-x_n(1),x_n(2))$ approximates $p_n(x_n)$ with relative error {\em exponentially decaying} in $n$ for $x_n = \lfloor nx \rfloor$, $x \in {\mathbb R}_+^2$, $0 < x(1) + x(2) < 1$. The steps of our analysis: 1) with an affine transformation, move the origin $(0,0)$ to $(n,0)$ on $\partial A_n$; let $n\nearrow \infty$ to remove the constraint on the $x(2)$ axis; this step gives the limit {\em unstable} /{\em transient} constrained random walk $Y$ and reduces $P_{x}(\tau_n < \tau_0)$ to $P_y(\tau < \infty)$; 2) construct a basis of harmonic functions of $Y$ and use it to apply the superposition principle to compute $P_y(\tau < \infty).$ The construction involves the use of conjugate points on a characteristic surface associated with the walk $X$. The proof that the relative error decays exponentially uses a sequence of subsolutions of a related HJB equation on a manifold.]]></description><subject>Mathematics - Probability</subject><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2018</creationdate><recordtype>article</recordtype><sourceid>GOX</sourceid><recordid>eNotj8tOwzAURL1hURU-gBX-gQQ7tuNkWarwkCoBUvbRtX2NLNI6stsS_p62dDWb0Zw5hNxzVspGKfYIaQ7HkjeMl0zWWi5It5qmFOewhX2IOxo97WaLOYcj0iew32P8oh8pGjBhDPuA-VzpfyLtYedwSz8PeMB8S248jBnvrrkk_XPXr1-LzfvL23q1KeAEK5pWemO1do2QrffgUHohhFO8Et7LyqK3BlrnFKsFoJFa16qqFDht0XAtluThf_biMUzpdDv9Dmef4eIj_gABZ0aW</recordid><startdate>20180115</startdate><enddate>20180115</enddate><creator>Sezer, Ali Devin</creator><scope>AKZ</scope><scope>GOX</scope></search><sort><creationdate>20180115</creationdate><title>Approximation of Excessive Backlog Probabilities of Two Tandem Queues</title><author>Sezer, Ali Devin</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-a674-894fbc77d8349ffade4f333d5123ff42cefcba9dd5063aeb47765225ad7ceb173</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2018</creationdate><topic>Mathematics - Probability</topic><toplevel>online_resources</toplevel><creatorcontrib>Sezer, Ali Devin</creatorcontrib><collection>arXiv Mathematics</collection><collection>arXiv.org</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Sezer, Ali Devin</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Approximation of Excessive Backlog Probabilities of Two Tandem Queues</atitle><date>2018-01-15</date><risdate>2018</risdate><abstract><![CDATA[Let $X$ be the constrained random walk on ${\mathbb Z}_+^2$ taking the steps $(1,0)$, $(-1,1)$ and $(0,-1)$ with probabilities $\lambda < (\mu_1\neq \mu_2)$; in particular, $X$ is assumed stable. Let $\tau_n$ be the first time $X$ hits $\partial A_n = \{x:x(1)+x(2) = n \}$ For $x \in {\mathbb Z}_+^2, x(1) + x(2) < n$, the probability $p_n(x)= P_x( \tau_n < \tau_0)$ is a key performance measure for the queueing system represented by $X$. Let $Y$ be the constrained random walk on ${\mathbb Z} \times {\mathbb Z}_+$ with increments $(-1,0)$, $(1,1)$ and $(0,-1)$. Let $\tau$ be the first time that the components of $Y$ equal each other. We derive the following explicit formula for $P_y(\tau < \infty)$: \[ P_y(\tau < \infty) = W(y)= \rho_2^{y(1)-y(2)} + \frac{\mu_2 - \lambda}{\mu_2 - \mu_1} \rho_1^{ y(1)-y(2)} \rho_1^{y(2)} + \frac{\mu_2-\lambda}{\mu_1 -\mu_2} \rho_2^{y(1)-y(2)} \rho_1^{y(2)}, \] where, $\rho_i = \lambda/\mu_i$, $i=1,2$, $y \in {\mathbb Z}\times{ \mathbb Z}_+$, $y(1) > y(2)$, and show that $W(n-x_n(1),x_n(2))$ approximates $p_n(x_n)$ with relative error {\em exponentially decaying} in $n$ for $x_n = \lfloor nx \rfloor$, $x \in {\mathbb R}_+^2$, $0 < x(1) + x(2) < 1$. The steps of our analysis: 1) with an affine transformation, move the origin $(0,0)$ to $(n,0)$ on $\partial A_n$; let $n\nearrow \infty$ to remove the constraint on the $x(2)$ axis; this step gives the limit {\em unstable} /{\em transient} constrained random walk $Y$ and reduces $P_{x}(\tau_n < \tau_0)$ to $P_y(\tau < \infty)$; 2) construct a basis of harmonic functions of $Y$ and use it to apply the superposition principle to compute $P_y(\tau < \infty).$ The construction involves the use of conjugate points on a characteristic surface associated with the walk $X$. The proof that the relative error decays exponentially uses a sequence of subsolutions of a related HJB equation on a manifold.]]></abstract><doi>10.48550/arxiv.1801.04674</doi><oa>free_for_read</oa></addata></record>
fulltext fulltext_linktorsrc
identifier DOI: 10.48550/arxiv.1801.04674
ispartof
issn
language eng
recordid cdi_arxiv_primary_1801_04674
source arXiv.org
subjects Mathematics - Probability
title Approximation of Excessive Backlog Probabilities of Two Tandem Queues
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-18T08%3A53%3A07IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-arxiv_GOX&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Approximation%20of%20Excessive%20Backlog%20Probabilities%20of%20Two%20Tandem%20Queues&rft.au=Sezer,%20Ali%20Devin&rft.date=2018-01-15&rft_id=info:doi/10.48550/arxiv.1801.04674&rft_dat=%3Carxiv_GOX%3E1801_04674%3C/arxiv_GOX%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rfr_iscdi=true