We study the problem of maintaining the distances and the shortest paths from a source node in a directed graph with arbitrary arc weights, when weight updates of arcs are performed. We propose algorithms that work for any graph and require linear space and optimal query time. If a negative-length cycle is added during weight-decreaseop-erations it is detected by the algorithms. The algorithms explicitly deal with zero-length cycles. We show that, if the graph has a k-bounded accounting function (as in the case of graphs with genus, arboricity, degree, tree width or pagenumber bounded by k, and k-inductive graphs) the algorithms require O(k · n · logn) worst case time. In the case of graphs with n nodes and m arcs k = O(√m); this gives O(√m · n · logn) worst case time per operation, which is better for a factor of O(√mlogn) than recomputing everything from scratch after each update. If we perform also insertions and deletions of arcs, then the above bounds become amortized. © Springer-Verlag Berlin Heidelberg 1998.
CITATION STYLE
Frigioni, D., Marchetti-Spaccamela, A., & Nanni, U. (1998). Fully dynamic shortest paths and negative cycles detection on digraphs with arbitrary arc weights. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 1461 LNCS, pp. 320–331). Springer Verlag. https://doi.org/10.1007/3-540-68530-8_27
Mendeley helps you to discover research relevant for your work.