On Traffic Engineering with Segment Routing in SDN based WANs

Segment routing is an emerging technology to simplify traffic engineering implementation in WANs. It expresses an end-to-end logical path as a sequence of segments, each of which is represented by a middlepoint. In this paper, we arguably conduct the first systematic study of traffic engineering with segment routing in SDN based WANs. We first provide a theoretical characterization of the problem. We show that for general segment routing, where flows can take any path that goes through a middlepoint, the resulting traffic engineering is NP-hard. We then consider segment routing with shortest paths only, and prove that the traffic engineering problem can now be solved in (weakly) polynomial time when the number of middlepoints per path is fixed and not part of the input. Our results thus explain, for the first time, the underlying reason why existing work only focuses on segment routing with shortest paths. In the second part of the paper, we study practical traffic engineering using shortest path based segment routing. We note that existing methods work by taking each node as a candidate middlepoint. This requires solving a large-scale linear program which is prohibitively slow. We thus propose to select just a few important nodes as middlepoints for all traffic. We use node centrality concepts from graph theory, notably group shortest path centrality, for middlepoint selection. Our performance evaluation using realistic topologies and traffic traces shows that a small percentage of the most central nodes can achieve good results with orders of magnitude lower runtime.