Data mining for diagnostic debugging in sensor networks: Preliminary evidence and lessons learned

Tarek Abdelzaher, Mohammad Khan, Hieu Le, Hossein Ahmadi, Jiawei Han

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Sensor networks and pervasive computing systems intimately combine computation, communication and interactions with the physical world, thus increasing the complexity of the development effort, violating communication protocol layering, and making traditional network diagnostics and debugging less effective at catching problems. Tighter coupling between communication, computation, and interaction with the physical world is likely to be an increasing trend in emerging edge networks and pervasive systems. This paper reviews recent tools developed by the authors to understand the root causes of complex interaction bugs in edge network systems that combine computation, communication and sensing. We concern ourselves with automated failure diagnosis in the face of non-reproducible behavior, high interactive complexity, and resource constraints. Several examples are given to finding bugs in real sensor network code using the tools developed, demonstrating the efficacy of the approach.

Original languageEnglish (US)
Title of host publicationKnowledge Discovery from Sensor Data - Second International Workshop, Sensor-KDD 2008, Revised Selected Papers
Pages1-24
Number of pages24
DOIs
StatePublished - Oct 20 2010
Event2nd International Workshop on Knowledge Discovery from Sensor Data, Sensor-KDD 2008 - Las Vegas, NV, United States
Duration: Aug 24 2008Aug 27 2008

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume5840 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other2nd International Workshop on Knowledge Discovery from Sensor Data, Sensor-KDD 2008
Country/TerritoryUnited States
CityLas Vegas, NV
Period8/24/088/27/08

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint

Dive into the research topics of 'Data mining for diagnostic debugging in sensor networks: Preliminary evidence and lessons learned'. Together they form a unique fingerprint.

Cite this