Joint Cross-Layer Optimization in Real-Time Networked Control Systems

Networked control system (NCS) refers to a set of control loops that are closed over a communication network. In this article, the joint operation of control and networking for NCS is investigated wherein the network serves the sensor-to-controller communication links for multiple stochastic linear time-invariant (LTI) subsystems. The sensors sample packets based on the observed plant state, which they send over a shared multihop network. The network has limited communication resources, which need to be assigned to competing links to support proper control loop operation. In this setup, we formulate an optimization problem to minimize the weighted-sum linear-quadratic-Gaussian cost of all loops, taking into account the admissible sampling, control, congestion control (CC), and scheduling policies. Under some mild assumptions on the sampling frequencies of the control loops and the communication network, we find the joint optimal solution to be given by a certainty equivalence control with a threshold-based sampling policy, as well as a back-pressure-type scheduler with a simple pass-through CC. The interface between network and control loops is identified to be the buffer state of the sensor node, which can be interpreted as network price for sampling a packet from the control perspective. We validate our theoretical claims by simulating NCSs comprised of multiple LTI stochastic control loops communicating over a two-hop cellular network.