Multi-Agent Adversarial Attacks for Multi-Channel Communications
This work addresses adversarial attacks in multi-channel communications, which is a practical but understudied domain, though it appears incremental as it extends existing RL-based adversarial methods to a multi-agent and multi-channel setting.
The paper tackles the problem of modeling adversaries in multi-channel wireless communication networks by proposing a multi-agent adversary system (MAAS) that learns to select channels and allocate power without prior knowledge of sender strategies, achieving a significant reduction in signal-to-noise ratio (SINR) compared to single-agent approaches under the same constraints.
Recently Reinforcement Learning (RL) has been applied as an anti-adversarial remedy in wireless communication networks. However, studying the RL-based approaches from the adversary's perspective has received little attention. Additionally, RL-based approaches in an anti-adversary or adversarial paradigm mostly consider single-channel communication (either channel selection or single channel power control), while multi-channel communication is more common in practice. In this paper, we propose a multi-agent adversary system (MAAS) for modeling and analyzing adversaries in a wireless communication scenario by careful design of the reward function under realistic communication scenarios. In particular, by modeling the adversaries as learning agents, we show that the proposed MAAS is able to successfully choose the transmitted channel(s) and their respective allocated power(s) without any prior knowledge of the sender strategy. Compared to the single-agent adversary (SAA), multi-agents in MAAS can achieve significant reduction in signal-to-noise ratio (SINR) under the same power constraints and partial observability, while providing improved stability and a more efficient learning process. Moreover, through empirical studies we show that the results in simulation are close to the ones in communication in reality, a conclusion that is pivotal to the validity of performance of agents evaluated in simulations.