Worst-Case TCAM Rule Expansion.

Abstract

Designers of TCAMs (Ternary CAMs) for packet classification deal with unpredictable sets of rules, resulting in highly variable rule expansions, and rely on heuristic encoding algorithms with no reasonable expansion guarantees. In this paper, given several types of rules, we provide new upper bounds on the TCAM worst-case rule expansions. In particular, we prove that a W-bit range can be encoded using W TCAM entries, improving upon the previously-known bound of 2W-5. We also propose a modified TCAM architecture that uses additional logic to significantly reduce the rule expansions, both in the worst case and in experiments with real-life classification databases.