Optimal routing in shortest-path data networks

Abstract

In this paper we address the problem of optimal routing in shortest-path data networks — such as Internet protocol (IP) networks — using open shortest path first (OSPF), routing information protocol (RIP), and other interior gateway protocols (IGPs). We propose a combinatorial algorithm for solving the optimal shortest-path routing problem, which consists of finding the link-distance metric assignment to all links in a network that results in optimal shortest-path routing and minimizes the average delay of packets in the network. Minimizing the delay of packets is critical, especially in the emerging voice over IP (VoIP) networks, quality of service (QoS)-centric virtual private networks (VPNs), and networks carrying other delay-sensitive traffic flows. Finding optimal link metrics that result in shortest-path routing is a special case of a classical inverse shortest-path problem. We propose a combinatorial approximation algorithm and illustrate the effectiveness of the proposed algorithm by comparing the results with bounds for several data network instances. We investigate not only the sensitivity of the solution with respect to the starting point, but also the sensitivity of the algorithm's performance with respect to uncertainty in traffic. We show that one preferred starting point consistently converges to near-optimal solutions. In addition, we extract two other distance metrics from the solution to the optimal general routing problem and show that these are preferred starting points for small or sparsely connected networks.