I’m with SONIC lab, Fudan University.
My research interests generally include software defined networking and Internet measurement, with a focus on quality of service and scalability issues. I’m interested in both designing theoretical algorithms, and in developing practical systems that are deployable in the real Internet.
Software defined networking
- Control Plane Scalability
- Path Computation
- Flowtable Lookup Acceleration
OpenFlow has been widely used in Software Defined Networking (SDN) to customize data plane behaviors through the policies at a logically centralized controller.
The centralized control plane design brings the potential of simplifying network management, but it also raises scalability concern for large-scale networks. We aim to build a highly scalable and flexible OpenFlow control plane. We propose the design and implementation of cCluster, which leverages the parallelism of cluster to balance the control plane load. Compared with existing solutions like Onix, cCluster can achieve better scalability, and can enable flexible management.
The fundamental tasks of the control plane in Software Defined Networking (SDN) are to customize forwarding policies for the data plane and to provide global network view for applications. The logically centralized design of control plane brings the benefit of network programmability and promises to ease network management. However, it also increases efficiency concerns for large-scale networks. Our goal is to build a high performance SDN control plane using multiple controllers. Deviating from conventional wisdom, we propose the design and implementation of ParaCon, which resorts to parallel computing to speed up the path computation in SDN control plane. We also address the consistency problem
and synchronization overhead under the design. To the best of our knowledge, ParaCon is the first attempt that utilizes node parallelism in path computation for SDN control plane.
We investigate the acceleration of Software-based OpenFlow switches, equipped with commodity off-the-shelf hardware, for high-performance table matching. Particularly, due to the high flexibility and compatibility, software-based and SDN-compatible switches, such as OpenvSwitch, has been widely applied in several viable fields, like cloud services, future Internet architectures, and the network function virtualization (NFV). In
these switches, table matching is a critical function. Existing CPU-based solutions are suffering from a low performance. In our work, we leverage the power of GPUs to accelerate table matching in software-based OpenFlow switches. We propose GFlow, which can handle OpenFlow table matching in a parallel fashion.
More SDN work…
My work is supported in part by the following grants.
- Large-scale SDN testbed for service convergence, 863 Program, I, 2015-2017
- 4K UHD Video Transmission over Next-Generation Broadcasting network, Science and Technology Commission of Shanghai Municipality, Co-PI, 2015-2017
- ,Home network platform for Internet of Things, Science and Technology Commission of Shanghai Municipality, Co-PI, 2014-2016
- On the Schematic and Algorithmic Aspects of Network Coding in P2P Networks, NSFC, PI, 2009-2011
- Next-Generation Routing with Network Coding, Shanghai Science and Technology Commission of Shanghai Municipality, PI, 2008-2011
- On the Quality of Service Model for Network Coding in P2P Networks, Research Fund for the Doctoral Program of Higher Education, MOE, PI, 2009-2011
- Network Coding for P2P Live Streaming, Shanghai Educational Development Foundation, PI, 2007-2009
- Data Dissemination with Network Coding, 863 Program, Co-I, 2006-2008