Statistical inverse problems in active network tomography

Abstract

The analysis of computer and communication networks gives rise tosome interesting inverse problems. This paper is concerned with activenetwork tomography where the goal is to recover information aboutquality-of-service (QoS) parameters at the link level from aggregatedata measured on end-to-end network paths. The estimation and monitoringof QoS parameters, such as loss rates and delays, are of considerableinterest to network engineers and Internet service providers. Thepaper provides a review of the inverse problems and recent researchon inference for loss rates and delay distributions. Some new resultson parametric inference for delay distributions are also developed.In addition, a real application on Internet telephony is discussed.