Resource Management in Softwarized Networks

Abstract

Communication networks are undergoing a major transformation through softwarization, which is changing the way networks are designed, operated, and managed. Network Softwarization is an emerging paradigm where software controls the treatment of network flows, adds value to these flows by software processing, and orchestrates the on-demand creation of customized networks to meet the needs of customer applications. Software-Defined Networking (SDN), Network Function Virtualization (NFV), and Network Virtualization are three cornerstones of the overall transformation trend toward network softwarization. Together, they are empowering network operators to accelerate time-to-market for new services, diversify the supply chain for networking hardware and software, bringing the benefits of agility, economies of scale, and flexibility of cloud computing to networks. The enhanced programmability enabled by softwarization creates unique opportunities for adapting network resources in support of applications and users with diverse requirements. To effectively leverage the flexibility provided by softwarization and realize its full potential, it is of paramount importance to devise proper mechanisms for allocating resources to different applications and users and for monitoring their usage over time.

The overarching goal of this dissertation is to advance state-of-the-art in how resources are allocated and monitored and build the foundation for effective resource management in softwarized networks. Specifically, we address four resource management challenges in three key enablers of network softwarization, namely SDN, NFV, and network virtualization. First, we challenge the current practice of realizing network services with monolithic software network functions and propose a microservice-based disaggregated architecture enabling finer-grained resource allocation and scaling. Then, we devise optimal solutions and scalable heuristics for establishing virtual networks with guaranteed bandwidth and guaranteed survivability against failure on multi-layer IP-over-Optical and single-layer IP substrate network, respectively. Finally, we propose adaptive sampling mechanisms for balancing the overhead of softwarized network monitoring and the accuracy of the network view constructed from monitoring data.