Abstract
Although adverse effects of attacks have been acknowledged in many cyber-physical systems, there is no rigorous mathematical analysis to characterize their worst effects in distributed multi-agent systems. Without characterizing these attacks, one cannot empower the agents with resilient functionalities to mitigate them. To this end, we will take the role of the attacker to show that in a distributed control system, an attacker can destabilize the whole synchronization process by injecting a state-independent attack signal into sensors or actuators of a single root node or to its outgoing communication links. This will be called the internal model principle for the attacker and will intensify the urgency of designing novel control protocols to mitigate these types of attacks.