Research Themes and Descriptions

This sub-group discusses challenges and research topics related to Software-defined networking (SDN) and netowrk function virtualization (NFV).

Fifth generation (5G) networks are expected to accommodate a wide range of services/use cases with diverse quality-of-service (QoS) requirements, such as machine-to-machine communication, Internet-of-Things (IoT), vehicular communications, and smart grid. In addition, future wireless and wired networks exhibit great heterogeneity in network entities and resources. Software-defined networking (SDN) and netowrk function virtulization (NFV) have a great potential to support diverse services/use cases by efficiently utilizing heterogeneous network resources in a flexible and agile manner. As an emerging and open paradigm, SDN decouples and migrates control functions from physical substrate networks to a logically centralized SDN controller, which simplifies the data plane packet forwarding, enables dynamic network management, and supports network programmability. Moreover, NFV virtualizes different network/service functions as software instances that are flexibily placed/programmed at different network servers, through which different sets of virtual network functions with associated resources create different network slices for various services. This process is referred to as network slicing. Network slices can be created and customized for different services/use cases to meet diversified quality-of-service (QoS) requirements. This research group aims to pave the road towards SDN/NFV-enabled 5G networks, by investigating the following key issues for SDN and NFV: SDN/NFV-enabled virtual network topology design, protocol customization, resource allocation, network embedding, and network slicing. We also investigate applying SDN and NFV technologies/architectures to other newly emerging networking/application scenarios in 5G/5G+ networks, including space-air-ground communication networks, UAV/drone-assisted networks, mobile edge/fog computing (MEC)-enabled computing resource slicing/offloading, artificial intelligence (AI) and machine learning (ML) based network deployment and reource allocation.