Using Models for Communication in Cyber-Physical Systems

Abstract

One of the main components of cyber-physical systems (CPS) is the underlying communication mechanism that enables control and decision-making. Communication has traditionally taken the form of sensing a physical phenomenon, or a cyber process, and then transmitting the sensed data to other entities within the system. With CPS being in general much more complex than a single physical or cyber process, the requirements on communication and data content are high. Therefore, communication of all the required information for control of a CPS may become a challenge. In this chapter, we present a new paradigm in communication which utilizes communication of models and model updates rather than raw sensed data. This approach, which transforms overall communication structure, has the potential to considerably reduce the communication load, and provide a mechanism for richer understanding of the processes whose data is being received over a communication link. We take the example of a vehicular CPS that relies on communication for collision avoidance and demonstrate the effectiveness of the model-based communication (MBC) concept.