Show simple item record

dc.contributor.authorBallapuram, Vijayanand Sreenivasanen
dc.date.accessioned2014-03-14T20:37:45Zen
dc.date.available2014-03-14T20:37:45Zen
dc.date.issued2007-04-16en
dc.identifier.otheretd-05202007-143854en
dc.identifier.urihttp://hdl.handle.net/10919/33033en
dc.description.abstractLand mobile radio (LMR) systems are used for communication by public safety and other government and commercial organizations. LMR systems offer mission-critical or even life-critical service in the day-to-day activities of such organizations. Traditionally, a variety of different LMR systems have been deployed by different organizations, leading to a lack of radio interoperability. A voice application that connects LMR systems via a packet-switched network is called an LMR Voice over IP (LMRVoIP) system and is a potential solution to the interoperability problem. LMRVoIP systems are time critical, i.e., are delay and jitter sensitive. Transmission of LMRVoIP traffic in a congested packet-switched network with no quality of service (QoS) or priority mechanisms in place could lead to high delays and extreme variations in delay, i.e., high jitter, thus resulting in poor application performance. LMRVoIP systems may also have performance issues with the use of virtual private networks (VPNs). To the best of our knowledge, there has been no prior thorough investigation of the performance of an LMRVoIP system with different queuing schemes for QoS and with the use of VPN. In this thesis, we investigate the performance of an LMRVoIP system with different queuing schemes and with the use of VPN. An experimental test bed was created to evaluate four QoS queuing schemes: first-in first-out queuing (FIFO), priority queuing (PQ), weighted fair queuing (WFQ), and class-based weighted fair queuing (CBWFQ). Quantitative results were obtained for voice application throughput, delay, jitter, and signaling overhead. Results show that, compared to a baseline case with no background traffic, LMRVoIP traffic suffers when carried over links with heavy contention from other traffic sources when FIFO queuing is used. There is significant packet loss for voice and control traffic and jitter increases. FIFO queuing provides no QoS and, therefore, should not be used for critical applications where the network may be congested. The situation can be greatly improved by using one of the other queuing schemes, PQ, WFQ, or CBWFQ, which perform almost equally well with one voice flow. Although PQ has the best overall performance, it tends to starve the background traffic. CBWFQ was found to have some performance benefits over WFQ in most cases and, thus, is a good candidate for deployment. The LMRVoIP application was also tested using a VPN, which led to a modest increase in latency and bandwidth utilization, but was found to perform well.en
dc.publisherVirginia Techen
dc.relation.haspartVijayBallapuramThesis.pdfen
dc.relation.haspartVijayBallapuramThesisv2.pdfen
dc.rightsIn Copyrighten
dc.rights.urihttp://rightsstatements.org/vocab/InC/1.0/en
dc.subjectQuality of Serviceen
dc.subjectVoice Over IPen
dc.subjectLand Mobile Radioen
dc.subjectVirtual Private Networksen
dc.titleImpact of Queuing Schemes and VPN on the Performance of a Land Mobile Radio VoIP Systemen
dc.typeThesisen
dc.contributor.departmentElectrical and Computer Engineeringen
dc.description.degreeMaster of Scienceen
thesis.degree.nameMaster of Scienceen
thesis.degree.levelmastersen
thesis.degree.grantorVirginia Polytechnic Institute and State Universityen
thesis.degree.disciplineElectrical and Computer Engineeringen
dc.contributor.committeechairMidkiff, Scott F.en
dc.contributor.committeememberDaSilva, Luiz A.en
dc.contributor.committeememberPark, Jung-Min Jerryen
dc.identifier.sourceurlhttp://scholar.lib.vt.edu/theses/available/etd-05202007-143854/en
dc.date.sdate2007-05-20en
dc.date.rdate2012-04-30en
dc.date.adate2007-07-23en


Files in this item

Thumbnail
Thumbnail

This item appears in the following Collection(s)

Show simple item record