Control-aware Uplink Resource Allocation for Cyber-Physical Systems in Wireless Networks

Abstract

Cellular networks beyond LTE must address the requirements of Machine Type Communications (MTC). Many MTC applications fall into category of the cyber-physical systems, i.e., feedback systems with the physical processes integrated with networks and computation. The challenge of CPS is to design the communication network to support the underlying feedback control loop. To address this challenge, we adopt a cross-layer approach to scheduling in wireless networks. We formulate the resource allocation problem with the objective of maximizing the control performance in terms of the network-induced error. Following that, we devise a Maximum Predicted Error First (MPEF) scheduler, which provides a close to optimal performance while only relying on offline information about the control loops. In a case study of LTE cell, we compare MPEF with state-ofthe-art scheduling algorithms.