An Analysis of Rule-set Databases in Packet Classification

Packet classification has proved to be an important challenge in network processing. It requires comparing each packet against a database of rules and forwarding the packet according to the highest priority matching rule. Packet classification can be seen as the categorization of incoming packets based on their headers according to specific criteria that examine specific fields within a packet header. The criteria are comprised of a set of rules that specify the content of specific packet header fields to result in a match. A packet classifier can be implemented in either software or hardware. An important entity in packet classification algorithms encompasses the rule-set database which includes a set of rules. Each rule is comprised of different fields such as source-IP, destination-IP, source-port, destination-port and protocol fields. A comprehensive and rigorous analysis of these fields appears to be the first systematic step to tackle the problem of packet classification and probably some elaborate optimizations for the existing algorithms. In this paper, we carefully analyze different rule-set databases and present the processed data with illustrative diagrams regarding the distributions of source/destination IPs, source/destination ports and protocol fields in the rule-sets. Based on the extracted information from these databases, we also investigate the outline of hash-based packet classification algorithms and present some trends to optimize this category of packet classification algorithms.