Software defined autonomic QoS model for future Internet

Abstract

Software defined technology has gained enormous momentum in both industry and academia. It may change the existing information flow architecture, which centered at hardware, by granting more privileges to deploy third-party software with more flexible network control and management capabilities. However, how to provide intelligent QoS guarantee in future network is still a hot research issue. Based on top-down design strategy, in this paper, we propose a hierarchical autonomic QoS model by adopting software-defined technologies. In this model, the network is divided into: data-plane, control-plane and management-plane. Then the data-plane is sliced as multiple virtual networks with the dynamic resource allocation capabilities, in order to carry different services or applications. The control-plane is abstracted as resource control layer and service control layer for each virtual network respectively. The management-plane then constructs hierarchical autonomic control loops in terms of different services’ bearing virtual networks and the corresponding hierarchical control logic. With the hierarchical autonomic control loop and layered service management functions, this model can provide dynamic QoS guarantee to the underlying data networks. We further proposed a context-aware Collaborative Borrowing Based Packet-Marking (CBBPM) algorithm. As an important underlying QoS component in data-plane, CBBPM can be combined with other QoS components to provide some new QoS functions. Experiment comparisons between different QoS mechanisms demonstrate the effectiveness of our proposed hierarchical autonomic QoS model and the packet marking mechanism.