Extension of the OpenFlow framework to foster the Software Defined Network paradigm in the Future Internet

Abstract: Software Defined Network (SDN) is one of the most promising paradigm for the Future Internet (FI), as it is a network architectural approach that enables the network virtualization and its centralized control with open interfaces for network applications and services. OpenFlow constitutes one of the pillar technologies for the development of the SDN, defining a protocol to control open switches and routers. Based on OpenFlow, several other research results have been achieved so that now the OpenFlow framework can be used to virtualize and program open networks. The aim of this work is to contribute to the development of the OpenFlow framework, by means of a proposal for the extension of the APIs made available to program and control the network. In particular, a novel network controller component, called ‘Traffic Statistics’, is presented in this work, able to collect network traffic information, to be used by cognitive resource management algorithms aiming at exploiting the network in a efficient way and at the same time at providing an adequate level of QoS to network flows. The proposed ‘Traffic Statistics ’ system design is presented here, together with the description of its preliminary prototypal implementation. 1