Research Article Evaluation of Spatial Parallel Genetic Algorithms for Real Time Routing in Geographi c I nformation System R. Shad , A. Shad , D. Molaei and M.S. Mesgari ABSTRACT In this study, two developed shortest path algorithms that run fast on the real large volume road networks have been identified. The first one is heuristic genetic algorithm implemented with approximate buckets in scalar computing environment and the second one is parallel genetic processing which is run in the alternative space. At first, these two algorithms were reviewed and summarized and their data structures and procedures are presented. Continually, in this effort genetic algorithm is used to solve the shortest path problem, because the limitation of traditional optimization methods. Finally, present result demonstrates that parallel heuristic processing can produce better speed up performance for real-time transportation applications. Services E-mail This Article Related Articles in ASCI Similar Articles in this Jou Search in Google Scholar View Citation Report Citation How to cite this article: R. Shad, A. Shad, D. Molaei and M.S. Mesgari, 2009. Evaluation of Spatial Parallel Genetic Algorithms for Real Time Routing Geographic Information System. Journal of Applied Sciences, 9: 4179-4182. DOI: 10.3923/jas.2009.4179.4182 URL: http://scialert.net/abstract/?doi=jas.2009.4179.4182 I NTRODUCTI ON Real-time transportation applications commonly require real-time processing while demanding best solutions in GIS environme example, today for the police applications or emergency services it is possible to find the fastest route and dispatch agents usin Recently, optimum path problems are discussed in computational geometry, graph algorithms, geographical information and ro Over many years researchers faced the problem of routing to (Maheshwari et al ., 2000; Mitchell, 2000) achieve optimum solut fast performance. Mitchell and Papadimitriou (1991) provide an approximation algorithm to compute a weighted short path. La al. (1997) described the cost of the approximation is no more than the shortest path cost plus a factor of Wj. Max-Planck Instit Computer Science is explained if ε is 1/100 then the number of Steinerpoints is reduced by a factor of 1/10. For two polyhedra with n nodes, Baltsan and Sharir (Baltsan and Sharir, 1988) presented an O(n 3 log n) time shortest path algorithm. Hershberge Suri (1995) introduced a linear time algorithm. Subsequently, multiple goals, factors and constraints in the large volume networks caused different heuristic and probabilistic m that no always guarantee the optimal solutions are presented. A number of these algorithms in the literatures are reported (Zh Noon, 1998; Goldberg and Radzik, 1993). Dial (1969) was the first one who implements the Dijkstra algorithm using buckets a heuristic method. Dial's original implementation (DKB) requires nC+1 buckets in the worst case, where C is the maximum arc l network (Ahuja et al ., 1993). Continually, metaheuristic Genetic algorithm (Diaz et al ., 1996) is developed as a robust, flexible adaptive tool with optimal designing and programming of networks. Genetic algorithm can be successfully performed in the non problems and also it is appropriate to face the noisy combinatorial solutions associated to the real networks. Evaluation of Spatial Parallel Genetic Algorithms for Real Time Routin... http://scialert.net/fulltext/?doi=jas.2009.4179.4182&org=11 1 of 5 2011/01/28 07:14 ظ. ب