Hardware-based Packet Classification Made Fast and Efficient

To achieve fast packet classification, a hardware-based scheme, cross-producting recurrence (CPR), based on a formerly proposed cross-producting scheme is proposed. This scheme simplifies the classification procedure and decrease the distinct combinations of fields by hierarchically decomposing the multi-dimensional space. In the new scheme, the multi-dimensional space is endowed with a hierarchical property which self-divides into several smaller subspaces, whereas the procedure of packet classification is translated into repeatedly searching for matching subspaces. The required storage of the proposed scheme is significantly reduced since the distinct fields of subspaces is controlled by a pre-defined configuration and can be much less than that of the filters. The experimental results demonstrate the effectiveness and scalability of the proposed scheme

[1]  George Varghese,et al.  Fast and scalable layer four switching , 1998, SIGCOMM '98.

[2]  T. V. Lakshman,et al.  High-speed policy-based packet forwarding using efficient multi-dimensional range matching , 1998, SIGCOMM '98.

[3]  Venkatachary Srinivasan,et al.  Packet classification using tuple space search , 1999, SIGCOMM '99.

[4]  Nick McKeown,et al.  Packet classification on multiple fields , 1999, SIGCOMM '99.

[5]  George Varghese,et al.  Packet classification for core routers: is there an alternative to CAMs? , 2003, IEEE INFOCOM 2003. Twenty-second Annual Joint Conference of the IEEE Computer and Communications Societies (IEEE Cat. No.03CH37428).

[6]  George Varghese,et al.  Packet classification using multidimensional cutting , 2003, SIGCOMM '03.

[7]  Walid Dabbous,et al.  Survey and taxonomy of IP address lookup algorithms , 2001, IEEE Netw..