Signaling games in networked cyber-physical systems with strategic elements

This paper studies signaling games in cyber-physical systems with strategic components. The communication network of a cyber-physical system is modeled as a sensor network, which involves a single Gaussian state observed by many sensors, subject to additive independent Gaussian observation noises. The sensors communicate with the receiver over a coherent Gaussian multiple access channel. There are two groups of sensors-strategic and non-strategic. The common objective of the team of non-strategic sensors and the receiver is to reconstruct the underlying state with minimum mean squared error. The team of strategic sensors, on the other hand, strives to minimize a different distortion function, which depends on the state, the reconstruction at the receiver and the type (bias) variable-an independent random variable whose realization is available only to the strategic sensors. It is shown that the ability of the team of non-strategic sensors and the receiver to secretly agree on a random event, that is 'coordination', plays a key role in the analysis. The properties and scaling behavior of the Stackelberg equilibrium of this signaling game are analyzed, in conjunction with the set of affine communication strategies, depending on the aforementioned coordination capability.