Loading...

Computing the Shortest Path on Weighted Triangulated Irregular Networks by Multicore Algorithms

Ghayour Baghbani, Farzaneh | 2011

1098 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 42119 (19)
  4. University: Sharif University of Technology
  5. Department: Computer Engineering
  6. Advisor(s): Ghodsi, Mohammad
  7. Abstract:
  8. Shortest path computation is one the fundamental problems in computer science. Triangulated Irregular Networks (TINs) are used in computational geometry to represent terrians and geometric surfaces. One of the most efficient mothods to solve the shortest path problem on a TIN is reducing it to shortest path problem on a graph. This reduction from continuous space to discrete space results in approximate solutions, but acceptable in real applications. In real applications we still encounter a large graph and using the simple Dijkstra algorithm consumes a lot of times. Memory shortage is another issue. Parallel processing could be a solution in this case. Multicore industry caused a revoloution in parallel processing. Now we found multicore processors every where. But multicore processing is different from older distributed memory parallel system. Design and implementation of a efficient program for a multicore system needs much more effort. In this thesis we are trying to answer the shortest path queries on weighted TINs. On our approach we encounter single source shortest path problem on weighted graphs. We give a survey on sequential and parallel methods for speeding up shortest path queries on weighted graphs and suggest a new method based on shortcuts. The method is implemented by OpenMP and the experimental results on realistic data shows the efficiency of the algorithm
  9. Keywords:
  10. Shortest Path ; Multicore Processors ; Triangulated Irregular Network Terrian (TIN)

 Digital Object List

 Bookmark

...see more