Fisher information-based experiment design for network tomography

Ting He, Chang Liu, Ananthram Swami, Don Towsley, Theodoros Salonidis, Andrei Iu Bejan, Paul Yu

Research output: Contribution to journalConference article

5 Citations (Scopus)

Abstract

Network tomography aims to infer the individual performance of networked elements (e.g., links) using aggregate measurements on end-to-end paths. Previous work on network tomography focuses primarily on developing estimators using the given measurements, while the design of measurements is often neglected. We fill this gap by proposing a framework to design probing experiments with focus on probe allocation, and applying it to two concrete problems: packet loss tomography and packet delay variation (PDV) tomography. Based on the Fisher Information Matrix (FIM), we design the distribution of probes across paths to maximize the best accuracy of unbiased estimators, asymptotically achievable by the maximum likelihood estimator. We consider two widely-adopted objective functions: determinant of the inverse FIM (D-optimality) and trace of the inverse FIM (A-optimality). We also extend the A-optimal criterion to incorporate heterogeneity in link weights. Under certain conditions on the FIM, satisfied by both loss and PDV tomography, we derive explicit expressions for both objective functions. When the number of probing paths equals the number of links, these lead to closed-form solutions for the optimal design; when there are more paths, we develop a heuristic to select a subset of paths and optimally allocate probes within the subset. Observing the dependency of the optimal design on unknown parameters, we further propose an algorithm that iteratively updates the design based on parameter estimates, which converges to the design based on true parameters as the number of probes increases. Using packet-level simulations on real datasets, we verify that the proposed design effectively reduces estimation error compared with the common approach of uniformly distributing probes.

Original languageEnglish (US)
Pages (from-to)389-402
Number of pages14
JournalPerformance Evaluation Review
Volume43
Issue number1
DOIs
StatePublished - Jun 24 2015
EventACM SIGMETRICS International Conference on Measurement and Modeling of Computer Systems, SIGMETRICS 2015 - Portland, United States
Duration: Jun 15 2015Jun 19 2015

Fingerprint

Fisher information matrix
Tomography
Experiments
Packet loss
Error analysis
Maximum likelihood
Optimal design

All Science Journal Classification (ASJC) codes

  • Software
  • Hardware and Architecture
  • Computer Networks and Communications

Cite this

He, T., Liu, C., Swami, A., Towsley, D., Salonidis, T., Bejan, A. I., & Yu, P. (2015). Fisher information-based experiment design for network tomography. Performance Evaluation Review, 43(1), 389-402. https://doi.org/10.1145/2796314.2745862
He, Ting ; Liu, Chang ; Swami, Ananthram ; Towsley, Don ; Salonidis, Theodoros ; Bejan, Andrei Iu ; Yu, Paul. / Fisher information-based experiment design for network tomography. In: Performance Evaluation Review. 2015 ; Vol. 43, No. 1. pp. 389-402.
@article{27672f6ca7214789a0eb0cf11d7898ff,
title = "Fisher information-based experiment design for network tomography",
abstract = "Network tomography aims to infer the individual performance of networked elements (e.g., links) using aggregate measurements on end-to-end paths. Previous work on network tomography focuses primarily on developing estimators using the given measurements, while the design of measurements is often neglected. We fill this gap by proposing a framework to design probing experiments with focus on probe allocation, and applying it to two concrete problems: packet loss tomography and packet delay variation (PDV) tomography. Based on the Fisher Information Matrix (FIM), we design the distribution of probes across paths to maximize the best accuracy of unbiased estimators, asymptotically achievable by the maximum likelihood estimator. We consider two widely-adopted objective functions: determinant of the inverse FIM (D-optimality) and trace of the inverse FIM (A-optimality). We also extend the A-optimal criterion to incorporate heterogeneity in link weights. Under certain conditions on the FIM, satisfied by both loss and PDV tomography, we derive explicit expressions for both objective functions. When the number of probing paths equals the number of links, these lead to closed-form solutions for the optimal design; when there are more paths, we develop a heuristic to select a subset of paths and optimally allocate probes within the subset. Observing the dependency of the optimal design on unknown parameters, we further propose an algorithm that iteratively updates the design based on parameter estimates, which converges to the design based on true parameters as the number of probes increases. Using packet-level simulations on real datasets, we verify that the proposed design effectively reduces estimation error compared with the common approach of uniformly distributing probes.",
author = "Ting He and Chang Liu and Ananthram Swami and Don Towsley and Theodoros Salonidis and Bejan, {Andrei Iu} and Paul Yu",
year = "2015",
month = "6",
day = "24",
doi = "10.1145/2796314.2745862",
language = "English (US)",
volume = "43",
pages = "389--402",
journal = "Performance Evaluation Review",
issn = "0163-5999",
publisher = "Association for Computing Machinery (ACM)",
number = "1",

}

He, T, Liu, C, Swami, A, Towsley, D, Salonidis, T, Bejan, AI & Yu, P 2015, 'Fisher information-based experiment design for network tomography', Performance Evaluation Review, vol. 43, no. 1, pp. 389-402. https://doi.org/10.1145/2796314.2745862

Fisher information-based experiment design for network tomography. / He, Ting; Liu, Chang; Swami, Ananthram; Towsley, Don; Salonidis, Theodoros; Bejan, Andrei Iu; Yu, Paul.

In: Performance Evaluation Review, Vol. 43, No. 1, 24.06.2015, p. 389-402.

Research output: Contribution to journalConference article

TY - JOUR

T1 - Fisher information-based experiment design for network tomography

AU - He, Ting

AU - Liu, Chang

AU - Swami, Ananthram

AU - Towsley, Don

AU - Salonidis, Theodoros

AU - Bejan, Andrei Iu

AU - Yu, Paul

PY - 2015/6/24

Y1 - 2015/6/24

N2 - Network tomography aims to infer the individual performance of networked elements (e.g., links) using aggregate measurements on end-to-end paths. Previous work on network tomography focuses primarily on developing estimators using the given measurements, while the design of measurements is often neglected. We fill this gap by proposing a framework to design probing experiments with focus on probe allocation, and applying it to two concrete problems: packet loss tomography and packet delay variation (PDV) tomography. Based on the Fisher Information Matrix (FIM), we design the distribution of probes across paths to maximize the best accuracy of unbiased estimators, asymptotically achievable by the maximum likelihood estimator. We consider two widely-adopted objective functions: determinant of the inverse FIM (D-optimality) and trace of the inverse FIM (A-optimality). We also extend the A-optimal criterion to incorporate heterogeneity in link weights. Under certain conditions on the FIM, satisfied by both loss and PDV tomography, we derive explicit expressions for both objective functions. When the number of probing paths equals the number of links, these lead to closed-form solutions for the optimal design; when there are more paths, we develop a heuristic to select a subset of paths and optimally allocate probes within the subset. Observing the dependency of the optimal design on unknown parameters, we further propose an algorithm that iteratively updates the design based on parameter estimates, which converges to the design based on true parameters as the number of probes increases. Using packet-level simulations on real datasets, we verify that the proposed design effectively reduces estimation error compared with the common approach of uniformly distributing probes.

AB - Network tomography aims to infer the individual performance of networked elements (e.g., links) using aggregate measurements on end-to-end paths. Previous work on network tomography focuses primarily on developing estimators using the given measurements, while the design of measurements is often neglected. We fill this gap by proposing a framework to design probing experiments with focus on probe allocation, and applying it to two concrete problems: packet loss tomography and packet delay variation (PDV) tomography. Based on the Fisher Information Matrix (FIM), we design the distribution of probes across paths to maximize the best accuracy of unbiased estimators, asymptotically achievable by the maximum likelihood estimator. We consider two widely-adopted objective functions: determinant of the inverse FIM (D-optimality) and trace of the inverse FIM (A-optimality). We also extend the A-optimal criterion to incorporate heterogeneity in link weights. Under certain conditions on the FIM, satisfied by both loss and PDV tomography, we derive explicit expressions for both objective functions. When the number of probing paths equals the number of links, these lead to closed-form solutions for the optimal design; when there are more paths, we develop a heuristic to select a subset of paths and optimally allocate probes within the subset. Observing the dependency of the optimal design on unknown parameters, we further propose an algorithm that iteratively updates the design based on parameter estimates, which converges to the design based on true parameters as the number of probes increases. Using packet-level simulations on real datasets, we verify that the proposed design effectively reduces estimation error compared with the common approach of uniformly distributing probes.

UR - http://www.scopus.com/inward/record.url?scp=84955567506&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84955567506&partnerID=8YFLogxK

U2 - 10.1145/2796314.2745862

DO - 10.1145/2796314.2745862

M3 - Conference article

AN - SCOPUS:84955567506

VL - 43

SP - 389

EP - 402

JO - Performance Evaluation Review

JF - Performance Evaluation Review

SN - 0163-5999

IS - 1

ER -