Abstract | ||
---|---|---|
There are two main optimization problems in software defined networks' traffic engineering, selecting conventional nodes to upgrade to SDN nodes, and finding a routing technique that minimize link utilization. While the second problem is almost solved, there are still not many theoretical results for the first problem. The approximation ratio of the only approximation algorithm for these problems is still as large as O(n), when n is the number of nodes. In this paper, we show that the problem is NP-hard, even for the single-source single-sink case. We then develop an algorithm for the case, and prove that our algorithm is kd-approximation algorithm, when k is the number of nodes to upgrade and d is the maximum degree of networks. We strongly believe that our algorithm can be extended to general cases in future.
|
Year | DOI | Venue |
---|---|---|
2018 | 10.1145/3164541.3164600 | IMCOM |
Keywords | Field | DocType |
Networking/ Telecommunications, SDN, Graph Theory | Graph theory,Approximation algorithm,Computer science,Computer network,Upgrade,Degree (graph theory),Software-defined networking,Traffic engineering,Optimization problem,Computer engineering | Conference |
ISBN | Citations | PageRank |
978-1-4503-6385-3 | 0 | 0.34 |
References | Authors | |
13 | 2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Xusheng Zhang | 1 | 39 | 13.00 |
Vorapong Suppakitpaisarn | 2 | 28 | 13.11 |