Parallelizing Count-Min Sketch Algorithm on Multi-core Processors

DOI

10.2991/mmebc-16.2016.472How to use a DOI?

Keywords

Count-Min sketch, Parallel algorithms, Frequent items.

Abstract

In this paper, we present a novel method that exploits the great parallel capability of multi-cores to speed up the famous Count-Min sketch algorithm. The proposed parallel Count-Min sketch algorithm equally distributes the input data stream into sub-threads which use the original Count-Min sketch algorithm to process the sub-streams. The counters in each local Count-Min sketch with frequency increments exceeding a pre-defined threshold are sent to a merging thread which is able to return the estimated frequencies satisfying the ( , )-approximation requirement. Experiments with real traffic traces demonstrate the excellent performance as well as the effects of parameters. The parallel Count-Min sketch algorithm achieves near-linear speedup at the cost of greater memory use.

Copyright

© 2016, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Bowen Yu
AU  - Yu Zhang
AU  - Lubing Li
PY  - 2016/06
DA  - 2016/06
TI  - Parallelizing Count-Min Sketch Algorithm on Multi-core Processors
BT  - Proceedings of the 2016 6th International Conference on Machinery, Materials, Environment, Biotechnology and Computer
PB  - Atlantis Press
SP  - 2342
EP  - 2345
SN  - 2352-5401
UR  - https://doi.org/10.2991/mmebc-16.2016.472
DO  - 10.2991/mmebc-16.2016.472
ID  - Yu2016/06
ER  -