Grant: Collaborative Research and Development Grants

The Internet started out as a breeding ground for innovation. One of its main design goals was to allow researchers to try out new and disruptive ideas. But, along the way, its amazing success and popularity has caused this very important objective to be sidelined. Due to the existence of multiple stakeholders, often with conflicting goals, major alterations to the existing Internet are next to impossible. Network virtualization has recently been propounded as a means to fend off this ossification. The main idea is to accommodate multiple logical (virtual) networks on a single physical network. The owner of a virtual network is free to deploy any communication protocol or service. By allowing multiple heterogeneous networks to cohabit on a shared physical one, the goal of network virtualization is to provide flexibility, diversity, security and manageability. However, this comes at a cost. Several new challenges need to be addressed. Recent research, to this end, can at best be described as attempts to fix some existing, and mostly disjoint, problems of the current Internet, rather than a coordinated effort to build a complete Network Virtualization Environment (NVE). In this project, we focus on three aspects that are of central importance to the instantiation and operation of an NVE, i) scalability, ii) failure resilience and iii) virtual networks management in a multi-layer production network. Advances in these research avenues in collaboration with our industrial partner Huawei could have a lasting impact on the design principles and management paradigms of the future Internet.