Learning attack mechanisms in Wireless Sensor Networks using Markov Decision Processes

作者：

Highlights：

• Markov Decision Processes are ideal to model attacks in wireless networks.

• Reinforcement learning allows learning to attack unknown defense systems.

• Deep reinforcement learning yields quasi optimal attack results.

• Deep reinforcement learning gives a trade off between attack results and complexity.

摘要

•Markov Decision Processes are ideal to model attacks in wireless networks.•Reinforcement learning allows learning to attack unknown defense systems.•Deep reinforcement learning yields quasi optimal attack results.•Deep reinforcement learning gives a trade off between attack results and complexity.

论文关键词：Markov Decision Process,Cooperative Spectrum Sensing,Dynamic Programming,Q-learning,Deep learning

论文评审过程：Received 28 May 2018, Revised 29 October 2018, Accepted 6 January 2019, Available online 7 January 2019, Version of Record 12 January 2019.

论文官网地址：https://doi.org/10.1016/j.eswa.2019.01.023