An online framework for flow round trip time measurement

Abstract

With the advent of high speed links, online flow measurement for, e.g., flow round trip time (RTT), becomes difficult due to the enormous amount of requirements on computational resources. We address this problem by proposing the double-deletion bloom filter (DDBF) scheme, which alleviates potential hash collisions of a standard bloom filter by explicitly deleting used records and implicitly deleting out-of-date records. As a result, it accurately estimates the current TCP flow RTT distribution with limited memory space, even with the appearance of multipath routing and Syn flooding attacks. Theoretical analysis indicates that the DDBF scheme exhibits desirable accuracy with a constant and small amount of memory. We further validate our scheme using real traces and demonstrate significant memory-saving without degrading accuracy.