Searching the Shortest Pair of Edge-Disjoint Paths in a Communication Network. A Fuzzy Approach

Abstract

In this paper, we address the problem of finding the shortest pair of edge-disjoint paths between two nodes in a communication network. We use a new cost function named modified fuzzy normalized used bandwidth, which is described as a fuzzy triangular number, thus incorporating the uncertainty generated in calculating this magnitude in a real network. The proposed algorithm uses as a sub-algorithm an adaptation of a Modified Fuzzy Dijkstra algorithm applied in a type V mixed graph with arcs whose costs are negative triangular fuzzy numbers, which has been described in previous work. We prove its effectivity by simulating traffic close to overload with two types of communication sources: regular and priority sending of information. The addressed problem presents a considerable interest in contexts such as finance entities or government services, where privacy and security against external attacks have to be considered.