Network Function Virtualization (NFV) is a new paradigm for providing elastic network functions through flexible virtual network function (VNF) instances executed on industry- standard computing platforms exemplified by cloud datacenters. We design efficient auction mechanisms for two types of NFV market.
First, we study the dynamic market mechanism for the transaction of VNF service chains in the NFV market, combining the techniques of primal-dual approximation algorithm de- sign with Myerson’s characterization of truthful mechanisms. Second, we consider the recent Cloud Radio Access Network (C-RAN) paradigm, a canonical use case of NFV. We study auction mechanisms for efficiently sharing C-RAN resources among mobile operators lever- aging randomized rounding, in both offline and online scenarios. The auction mechanisms we design run efficiently in polynomial time, guarantee truthfulness, and achieve near-optimal social welfare. The C-RAN online algorithm achieves a competitive ratio of (1−ε). Extensive simulation studies verify the efficacy of our auction mechanisms.