ABSTRACT

in a realm of ever-increasing Internet connectivity, together with swelling computer security threats, security-cognizant network applications technology is gaining widespread popularity. Packet classifiers are extensively employed for numerous network applications in different types of network devices such as Firewalls and Router, among others. Appreciating the tangible performance of recommended packet classifiers is a prerequisite for both algorithm creators and consumers. However, this is occasionally challenging to accomplish. Each innovative algorithm published is assessed from diverse perceptions and is founded on different suppositions. Devoid of a mutual foundation, it is virtually impossible to compare different algorithms directly. In the interim, it too aids the system implementers to effortlessly pick the most suitable algorithm for their actual applications.
Electing an ineffectual algorithm for an application can invite major expenditures. This is particularly true for packet classification in network routers, as packet classification is fundamentally a tough problem and all current algorithms are constructed on specific heuristics and filter set characteristics. The performance of the packet classification subsystem is vital for the aggregate success of the network routers.
In this study, we have piloted an advanced exploration of the existing algorithms to provide a comparative evaluation of a number of known classification algorithms that have been considered for both software and hardware implementation. We have explained our earlier suggested DimCut packet classification algorithm, and related it with the BV, HiCuts and HyperCuts decision tree-based packet classification algorithms with the comparative evaluation analysis. This comparison has been carried out on implementations based on the same principles and design choices from different sources. Performance measurements have been obtained by feeding the implemented classifiers with a large number of random Rules and Packets in the same test scenario.