Communications of the ACM
Cybersecurity Technique Keeps Hackers Guessing
The research team used deep reinforcement learning to gradually shape the behavior of the algorithm.
Credit: Cyber Security Hub
Development Command's Army Research Laboratory (ARL) has designed a machine learning-based framework to augment the security of in-vehicle computer networks.
The DESOLATOR (deep reinforcement learning-based resource allocation and moving target defense deployment framework) is engineered to help an in-vehicle network identify the optimal Internet Protocol (IP) shuffling frequency and bandwidth allocation to enable effective, long-term moving target defense.
Explained ARL's Terrence Moore, "If you shuffle the IP addresses fast enough, then the information assigned to the IP quickly becomes lost, and the adversary has to look for it again."
ARL's Frederica Free-Nelson said the framework keeps uncertainty sufficiently high to defeat potential attackers without incurring excessive maintenance costs, and prevents performance slowdowns in high-priority areas of the network.
From U.S. Army DEVCOM Army Research Laboratory
View Full Article
Abstracts Copyright © 2021 SmithBucklin, Washington, DC, USA
No entries found