PROJECT TITLE :

TCAM-Based Multi-Match Packet Classification Using Multidimensional Rule Layering

ABSTRACT:

Ternary content addressable memory (TCAM) has superior performance for single-match packet classification however not the case for multi-match packet classification. The limitation is caused by TCAM design that reports only the first matching rule. To cope with the limitation, previous algorithms use additional TCAM entries or accesses, or both, to meet multi-match packet classification. These algorithms also reorder rules; therefore, a multi-match classifier primarily based on these algorithms cannot maintain performance for single-match packet classification. In different words, all matching rules must be yielded to determine the best priority matching rule. During this paper, we gift a TCAM-primarily based theme for multi-match packet classification without single-match penalty. Our scheme partitions a rule set based mostly on vary layering, that will be applied to achieve vary encoding. The rule partitioning generates rule subsets that satisfy that the principles in a very subset are mutually disjoint. Every rule is then tagged a bitmap for subset identification to meet multi-match packet classification. 2 approaches, loose coupling and tight coupling, are derived with completely different search procedures whereas incorporating vary encoding. Each approaches will maintain original rule order, but with completely different performance tradeoff. We additionally gift a refinement which uses all offered TCAM entries to boost the performance of multi-match packet classification. The experimental results show that combining vary encoding with multi-match packet classification has blessings of storage potency and speed superiority. The capability of supporting single-match packet classification additionally provides better flexibility of applying completely different packet actions.