n order to address network dynamics and highly varying requirements, flexibility has emerged as a key property for networks to cope with increasing dynamics and to be prepared for future demands. Softwarized networks including concepts such as Network Virtualization, Software Defined Networking and Network Function Virtualization promise flexibility. However, so far flexibility is mainly used as a qualitative advantage for a certain design choice where the meaning of flexibility is varying a lot in literature. To provide a better understanding of how to design flexible networks, we propose a definition for flexibility and present an approach for a quantitative measure of flexibility in softwarized networks. In our proposal, we refer to flexibility as the ability to support new requests, e.g., changes in the requirements or new traffic distributions, in a timely manner. We illustrate with use case studies for function placement and SDN resilience, how this measure can be used to evaluate and compare different network designs quantitatively. To address adaptation time in flexible networks, we further present approaches to speed up the execution of algorithms based on machine learning. Examples include virtual network embedding and function placement. With our proposed approach for the definition and evaluation of flexibility, we intend to stimulate the discussion towards a more quantitative analysis of softwarized networks and beyond.     «

n order to address network dynamics and highly varying requirements, flexibility has emerged as a key property for networks to cope with increasing dynamics and to be prepared for future demands. Softwarized networks including concepts such as Network Virtualization, Software Defined Networking and Network Function Virtualization promise flexibility. However, so far flexibility is mainly used as a qualitative advantage for a certain design choice where the meaning of flexibility is varying a lot...     »