Loading...

Local Approximations for Graph Problems in Large-Scale Distributed Networks

Dastangoo, Hamed | 2015

632 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 46625 (19)
  4. University: Sharif University of Technology
  5. Department: Computer Engineering
  6. Advisor(s): Izadi, Mohammad; Zarrabi-Zadeh, Hamid
  7. Abstract:
  8. Local algorithm is a distributed algorithm which runs in constant number of rounds and its running time is independent of network size. Local approximation is a local algorithm which presents an approximation for optimal solution. In large networks (whether be natural or virtual), locality is a central subject. In networks like Internet, human society, human brain, and etc, nodes are confined to local information since accessing the whole topology of their graph is impossibile or expensive. Concerning ever-growing large networks like ad hoc and wireless sensor networks, having local algorithms for the tasks in these networks are very interested and becomes a necessity. Although there are some positive results in achieving exact solutions for some prob- lems but most interested problems in the networks are not solvable exactly, therefore achieving approximations for them is a reasonable choice. As there are gaps between upper bounds and lower bounds for these problems, trying to close these upper bounds is a typical try. During two past decades the interest in local computations was always increasing, but still the community is far from answering the question asked more than two decades ago: ”What can be computed locally?”. In this thesis the locality and local approximability in large-scale networks is investigated. For the purpose of investigat- ing a specific problem, Minimum Dominating Set problem is chosen. This problem is used in many situations in distributed settings like clustering, scheduling, and etc . The local approximability of this problem is investigated
  9. Keywords:
  10. Distributed Algorithm ; Local Model ; Local Algorithm ; Minimum Dominating Set ; Local Approximation

 Digital Object List

 Bookmark

...see more