Abstract | ||
---|---|---|
This paper presents PPS, a system for locating occurrences of string keywords stored in the payload of packets using a programmable network ASIC. The PPS compiler first converts keywords into Deterministic Finite Automata (DFA) representations, and then maps the DFA into a sequence of forwarding tables in the switch pipeline. Our design leverages several hardware primitives (e.g., TCAM, hashing, parallel tables) to achieve high throughput. Our evaluation shows that PPS demonstrates significantly higher throughput and lower latency than string searches running on CPUs, GPUs, or FPGAs.
|
Year | DOI | Venue |
---|---|---|
2019 | 10.1145/3314148.3314356 | SOSR |
Keywords | Field | DocType |
Programmable switches,String searching | Computer science,Deterministic finite automaton,Parallel computing,Network packet,Field-programmable gate array,Compiler,Application-specific integrated circuit,Hash function,Throughput,Distributed computing,Payload | Conference |
ISBN | Citations | PageRank |
978-1-4503-6710-3 | 6 | 0.51 |
References | Authors | |
0 | 7 |
Name | Order | Citations | PageRank |
---|---|---|---|
Theo Jepsen | 1 | 28 | 3.72 |
Daniel Álvarez | 2 | 212 | 25.22 |
J. Nathan Foster | 3 | 2074 | 110.56 |
Changhoon Kim | 4 | 1716 | 121.18 |
Jeongkeun Lee | 5 | 1089 | 56.72 |
Masoud Moshref | 6 | 263 | 13.73 |
Robert Soulé | 7 | 456 | 28.26 |