Ultra-High Throughput Low-Power Packet Classification

Kennedy, Alan and Wang, Xiaojun (2014) Ultra-High Throughput Low-Power Packet Classification. IEEE Transactions on Very Large Scale Integration (VLSI) Systems , 22 (2).

[thumbnail of kennedy.pdf]
Download (941kB) | Preview


Packet classification is used by networking equipment to sort packets into flows by comparing their headers to a list of rules, with packets placed in the flow determined by the matched rule. A flow is used to decide a packet's priority and the manner in which it is processed. Packet classification is a difficult task due to the fact that all packets must be processed at wire speed and rulesets can contain tens of thousands of rules. The contribution of this paper is a hardware accelerator that can classify up to 433 million packets per second when using rulesets containing tens of thousands of rules with a peak power consumption of only 9.03 W when using a Stratix III field-programmable gate array (FPGA). The hardware accelerator uses a modified version of the HyperCuts packet classification algorithm, with a new pre-cutting process used to reduce the amount of memory needed to save the search structure for large rulesets so that it is small enough to fit in the on-chip memory of an FPGA. The modified algorithm also removes the need for floating point division to be performed when classifying a packet, allowing higher clock speeds and thus obtaining higher throughputs.

Item Type: Article
Uncontrolled Keywords: Hardware accelerator; high throughput; low power; packet classification; parallel processing.
Subjects: Computer Science
Research Centres: Other
Depositing User: Sean McGreal
Date Deposited: 17 Oct 2016 08:26
Last Modified: 17 Oct 2016 08:26
License: Creative Commons: Attribution-Noncommercial-Share Alike 4.0

Actions (login required)

View Item View Item


Downloads per month over past year