基于自适应超时计数布鲁姆过滤器的流量测量算法

针对流量测量中IP长流的检测问题，该文设计了计数布鲁姆过滤器(Count Bloom Filter, CBF)与超时布鲁姆过滤器(Timeout Bloom Filter, TBF)结合的长流检测机制。该机制动态调整布鲁姆过滤器中的超时时间，及时清理结束流，解决空间拥塞问题，从而可以适用于无结束标志IP长流检测。依据算法整体错误率与超时时间的分析，根据链路流到达强度与布鲁姆过滤器向量空间长度自适应动态调整超时时间，使得算法整体错误率保持最低。该算法的性能利用真实网络流量数据进行验证，结果表明，与现有算法相比，该算法的测量准确性更高。

An Adaptive Timeout Counter Bloom Filter Algorithm for Traffic Measurement

A novel mechanism combining Counting Bloom Filter (CBF) and Timeout Bloom Filter (TBF) is proposed, aiming at identifying IP long flow precisely. By adjusting the timeout dynamically and deleting end flows timely, the mechanism can solve the space congestion of Bloom filter and identify heavy hitters without normal end flag. The timeout and accuracy are analyzed. When adjusting the timeout dynamically according to the traffic arrival intensity and Bloom filter vector length, the mechanism can get minimum error. The experiments are conducted based on the real network trace. The results demonstrate that the proposed method is more accurate than the existing algorithms.