Correlation between complex network features and robustness in linux kernel modules

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

After studying the node degree distribution and average path length of the kernel module call graphs of 223 Linux kernels (V1.1.0 to V2.4.35), we have found that the call graphs of the file system, drivers, kernel, and memory management modules are scale-free small-world complex networks that exhibit strong clustering tendency. Using the random error case and attack case methods, we investigated the robustness of the modules by removing the nodes in their call graphs. We further proposed the Critical Removal Ratio to measure the robustness of these modules from the point view of complex networks. Finally, we showed correlations between robustness and the complex network features of kernel modules.

Original languageEnglish
Title of host publicationProceedings - 2017 Annual Conference on Software Analysis, Testing and Evolution, SATE 2017
EditorsXiaoyuan Xie, Wei Dong
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages80-89
Number of pages10
ISBN (Electronic)9781538636879
DOIs
StatePublished - 22 Nov 2017
Event2017 International Conference on Software Analysis, Testing and Evolution, SATE 2017 - Harbin, Heilongjiang, China
Duration: 3 Nov 20174 Nov 2017

Publication series

NameProceedings - 2017 Annual Conference on Software Analysis, Testing and Evolution, SATE 2017
Volume2017-January

Conference

Conference2017 International Conference on Software Analysis, Testing and Evolution, SATE 2017
Country/TerritoryChina
CityHarbin, Heilongjiang
Period3/11/174/11/17

Fingerprint

Dive into the research topics of 'Correlation between complex network features and robustness in linux kernel modules'. Together they form a unique fingerprint.

Cite this