Dynamic IGP Weight Optimization in IP Networks

Abstract

With the explosive growth of the Internet and the incredible development of network applications, the variation in traffic volume has become one of the most important problems faced by network operators. Designing a network using a single ``busy hour'' traffic matrix strains credibility due to the high volatility of traffic patterns. Thus, there is a need for efficient dynamic reconfiguration methods allowing to adapt resource utilization to prevailing traffic. In this paper, we focus on the problem of link weight optimization in IP networks where the traffic is routed along shortest paths according to the link metrics (OSPF and IS-IS-based networks). We propose an online approach to handle time-varying traffic matrices that relies on online traffic monitoring and updates link weights, and thus the routing paths, adaptively as some changes are observed.