On the capacity of multihop wireless networks: Fundamental limitations and efficient algorithms

Simple item page
dc.contributor.advisorRiedi, Rudolf H.en_US
dc.creatorKeshavarzHaddad, Alirezaen_US
dc.date.accessioned2018-12-03T18:32:11Zen_US
dc.date.available2018-12-03T18:32:11Zen_US
dc.date.issued2007en_US
dc.description.abstractWireless networks are rapidly becoming a major component of the modern communications infrastructure competing with wireline networking. The and of this thesis is to study the capacity, fundamental limitations, and the throughput of a number of protocols in multihop wireless networks. To this end, we introduce novel frameworks and powerful tools for analyzing throughput and the effect of various network protocols. In the first part of the thesis, we study broadcast in wireless networks in several complementary directions. We start by proposing three novel broadcast schemes for ad hoc networks which combine low overhead with robustness and are provably efficient. Next, we study the dynamics of broadcast flooding under a realistic MAC model in large wireless networks. We find conditions on survivability and coverage of flooding, and prove that flooding disseminates the information in almost constant speed. Then, we introduce and compute the broadcast capacity for multihop wireless networks. We show that the broadcast capacity is proportional to the wireless channel bandwidth. Moreover, we demonstrate that only the broadcasts schemes with efficient backbone can achieve a throughput close to the broadcast capacity. By applying these results, we are able to analytically evaluate the throughput and the effect of broadcast applications in arbitrary wireless networks. In the second part of the thesis, we generalize our study from broadcast to unicast and multicast in wireless networks. We introduce novel analytical tools for analyzing the capacity of arbitrary wireless networks. We provide some tight and sensitive bounds on the network capacity for unicast and nnilticast applications which depend on topology and traffic pattern of the network. Furthermore, we add another dimension to our work by allowing network coding among the nodes. We investigate the benefit of network coding for some popular scenarios in wireless networks, and demonstrate that the network coding gain is limited by a constant factor. We identify wireless channel interference, network topology, and traffic pattern as the determining parameters on the capacity of wireless networks. Finally, we discuss how the relevant results of this thesis can be used for design and development of wireless networks in some particular network scenarios.en_US
dc.format.extent245 ppen_US
dc.identifier.callnoTHESIS E.E. 2008 KESHAVARZHADDADen_US
dc.identifier.citationKeshavarzHaddad, Alireza. "On the capacity of multihop wireless networks: Fundamental limitations and efficient algorithms." (2007) Diss., Rice University. <a href="https://hdl.handle.net/1911/103648">https://hdl.handle.net/1911/103648</a>.en_US
dc.identifier.digital304804326en_US
dc.identifier.urihttps://hdl.handle.net/1911/103648en_US
dc.language.isoengen_US
dc.rightsCopyright is held by the author, unless otherwise indicated. Permission to reuse, publish, or reproduce the work beyond the bounds of fair use or other exemptions to copyright law must be obtained from the copyright holder.en_US
dc.subjectElectrical engineeringen_US
dc.subjectComputer scienceen_US
dc.subjectApplied sciencesen_US
dc.subjectBroadcast schemesen_US
dc.subjectMultihop wireless networksen_US
dc.subjectNetwork capacityen_US
dc.titleOn the capacity of multihop wireless networks: Fundamental limitations and efficient algorithmsen_US
dc.typeThesisen_US
dc.type.materialTexten_US
thesis.degree.departmentElectrical Engineeringen_US
thesis.degree.disciplineEngineeringen_US
thesis.degree.grantorRice Universityen_US
thesis.degree.levelDoctoralen_US
thesis.degree.nameDoctor of Philosophyen_US
Files
Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
304804326.pdf
Size:
6.93 MB
Format:
Adobe Portable Document Format
Download
Collections
Rice University Theses and Dissertations