Parallelizing the Weighted Lossy Counting Algorithm in High-speed Network Monitoring

With the ever increasing of network traffic and link bandwidth, parallel frequent item mining becomes more and more important in high-speed network traffic engineering and monitoring. The advent of multi-core processors calls for efficient parallel designs which can effectively utilize the parallelism of multi-cores. In this paper, we present a novel precision integrated frame work (PRIF) to parallelize the famous Lossy Counting algorithm in the context of multi-core processors. PRIF achieves near-linear speedup and is capable of continuously monitoring the frequent items all along. Experiments with three real traffic traces demonstrate the excellent performance of PRIF.