1 / 6

Ultra-High Throughput Low-Power Packet Classification

This paper presents an innovative approach to packet classification, leveraging decision trees and HyperCuts to achieve ultra-high throughput with low power consumption. We introduce a novel method to optimize the number of cuts in packet classification trees by focusing on dimensions with high variability. Through techniques such as node merging, rule overlap reduction, and region compaction, we enhance the efficiency of rule storage and processing. The methodology ensures a balanced tree structure while maintaining classification speed and accuracy, making it suitable for modern networking demands.

tricia
Télécharger la présentation

Ultra-High Throughput Low-Power Packet Classification

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Ultra-High Throughput Low-PowerPacket Classification Author: Alan Kennedy and Xiaojun Wang Accepted by IEEE Transactions on VLSI

  2. Overview • Decision Tree based Packet Classification • Based on HyperCuts

  3. HyperCuts Review • Dimension to Cut • The fields with more distinct number of range than mean • Number of Cut • Max cuts to node i ≤ spfac * sqrt (number of rules at i) • Actually Cutting • Combination with least maximum number of rule stored in a child • Heuristics • Node Merging • Rule Overlap • Pushing Common Rule Upward • Region Compaction

  4. Modification • Dimension to Cut • All fields are considered • Number of Cut • Larger the number of cut in root-node (seg. Table) • Optimization • Pushing Common Rule Upward (Too Complicate) • Modified Region Compaction (Next Page) • Rule Storage

  5. Region Compaction: HyperCuts

  6. Region Compaction: Modification • Necessary Information: • Number of Cut in each field • Number of Bit in each field

More Related