<?xml version="1.0" encoding="utf-8"?>
<export-example>
  <doc>
    <id>1389</id>
    <completedYear/>
    <publishedYear/>
    <thesisYearAccepted/>
    <language>eng</language>
    <pageFirst/>
    <pageLast/>
    <pageNumber/>
    <edition/>
    <issue/>
    <volume/>
    <type>reportzib</type>
    <publisherName/>
    <publisherPlace/>
    <creatingCorporation/>
    <contributingCorporation/>
    <belongsToBibliography>0</belongsToBibliography>
    <completedDate>2011-09-01</completedDate>
    <publishedDate>2011-09-01</publishedDate>
    <thesisDateAccepted>--</thesisDateAccepted>
    <title language="eng">Robust Network Design: Formulations, Valid Inequalities, and Computations</title>
    <abstract language="eng">Traffic in communication networks fluctuates heavily over time.&#13;
  Thus, to avoid capacity bottlenecks, operators highly overestimate&#13;
  the traffic volume during network planning. In this paper we&#13;
  consider telecommunication network design under traffic uncertainty,&#13;
  adapting the robust optimization approach of Bertsimas and Sim [2004]. We&#13;
  present three different mathematical formulations for this problem,&#13;
  provide valid inequalities, study the computational implications,&#13;
  and evaluate the realized robustness.&#13;
&#13;
  To enhance the performance of the mixed-integer programming solver&#13;
  we derive robust cutset inequalities generalizing their&#13;
  deterministic counterparts. Instead of a single cutset inequality&#13;
  for every network cut, we derive multiple valid&#13;
  inequalities by exploiting the extra variables available in the&#13;
  robust formulations. We show that these inequalities define facets&#13;
  under certain conditions and that they completely describe a projection&#13;
  of the robust cutset polyhedron if the cutset consists of a single edge.&#13;
&#13;
  For realistic networks and live traffic measurements we compare the&#13;
  formulations and report on the speed up by the valid inequalities.&#13;
  We study the "price of robustness" and evaluate the&#13;
  approach by analyzing the real network load. The results show that&#13;
  the robust optimization approach has the potential to support&#13;
  network planners better than present methods.</abstract>
    <identifier type="issn">1438-0064</identifier>
    <identifier type="serial">11-34</identifier>
    <identifier type="urn">urn:nbn:de:0297-zib-13890</identifier>
    <author>Arie M.C.A. Koster</author>
    <submitter>Christian Raack</submitter>
    <author>Manuel Kutschka</author>
    <author>Christian Raack</author>
    <series>
      <title>ZIB-Report</title>
      <number>11-34</number>
    </series>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>robust network design</value>
    </subject>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>cutset inequalities</value>
    </subject>
    <collection role="msc" number="90Bxx">Operations research and management science</collection>
    <collection role="msc" number="90B18">Communication networks [See also 68M10, 94A05]</collection>
    <collection role="institutes" number="optimization">Mathematical Optimization</collection>
    <collection role="projects" number="MATHEON-B3">MATHEON-B3</collection>
    <collection role="institutes" number="aopt">Applied Optimization</collection>
    <file>https://opus4.kobv.de/opus4-zib/files/1389/20110901-ZIB-Report-11-34.pdf</file>
  </doc>
</export-example>
