Data Injection Attack on Decentralized Optimization
Sissi Xiaoxiao Wu, Hoi-To Wai, Anna Scaglione, Angelia Nedic, Amir Leshem
Abstract
This paper studies the security aspect of gossip-based decentralized optimization algorithms for multi agent systems against data injection attacks. Our contributions are two-fold. First, we show that the popular distributed projected gradient method (by Nedić et al.) can be attacked by <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">coordinated insider</i> attacks, in which the attackers are able to steer the final state to a point of their choosing. Second, we propose a metric that can be computed locally by the trustworthy agents processing their own iterates and those of their neighboring agents. This metric can be used by the trustworthy agents to detect and localize the attackers. We conclude the paper by supporting our findings with numerical experiments.
BibTeX
@inproceedings{icassp2018_datainjectionatt,
title = {Data Injection Attack on Decentralized Optimization},
author = {Sissi Xiaoxiao Wu and Hoi-To Wai and Anna Scaglione and Angelia Nedic and Amir Leshem},
booktitle = {ICASSP 2018},
year = {2018}
}