Optimal packet classification applicable tothe OpenFlow context

A configurable packet classification architecture for Software-Defined Networking

“…However, currently, large real-life rule sets are available with no more than five fields for analysis and performance. The authors in [8] and [9] created a scenario with 12-field rules, which has not been practically applied, while there is no evidence of the validity of the 15-field rules used in [10].…”

“…EffiCuts reduces memory space by reducing the number of overlapped rules but with increased memory access time. The same algorithm was proposed in [8] based on OpenFlow [1]. The authors group rules according to their sizes for efficient storage.…”

“…The proposed circuit is able to operate 42.73 Gbps with 40 bytes per packet as minimum size, overcoming the OC-768 rate. This is not achieved in software-based systems such as [8] and [10]. Table V summarizes the synthesis results.…”

An parallelized deep packet inspection design in software defined network

“…Stimpfling T [18] says that due to the emergence of new standards such as OpenFlow, packet classification algorithms have to be reconsidered to support effectively classification over more than 5 fields and analyze the performance offered by EffiCuts in the context of OpenFlow [20]. Kogan K etc [22] identify methods from other fields of computer science and show research directions that can be applied to efficient design of packet classifiers and propose abstractions and design patterns can significantly reduce requirements on network elements and enable deployment of functionality that would be infeasible in a traditional way.…”