Measurement of loss pairs in network paths

Loss-pair measurement was proposed a decade ago for discovering network path properties, such as a router's buffer size. A packet pair is regarded as a loss pair if exactly one packet is lost. Therefore, the residual packet's delay can be used to infer the lost packet's delay. Despite this unique advantage shared by no other methods, no loss-pair measurement in actual networks has ever been reported. In this paper, we further develop the loss-pair measurement and make the following contributions. First, we characterize the residual packet's delay by including other important factors (such as the impact of the first packet in the pair) which were ignored before. Second, we employ a novel TCP-based probing method to measure from a single endpoint all four possible loss pairs for a round-trip network path. Third, we conducted loss-pair measurement for 88 round-trip paths continuously for almost three weeks. Being the first set of loss-pair measurement, we obtained a number of original results, such as prevalence of loss pairs, distribution of different types of loss pairs, and effect of route change on the paths' congestion state.