A space-saving method for aggregate Top-N flow statistics with high accuracy

In this paper, we propose a new space-saving method for aggregate Top-N flow statistics. Based on the heavy-tailed characteristic in flow statistics, we divide the flow data into two data sets, mainly focus on the non-redundant data set, and restrict the maximum size of redundant data set. By doing this, the total amount of storage space is reduced. To ensure the statistics accuracy, we use the Least Recently Updated elimination algorithm to keep the useful data and discard the data which matters less to the result. The experimental results show that our method has a high accuracy.