Resource-Optimal Heterogeneous Machine-to-Machine Communications in Software Defined Networking Cyber-Physical Systems

Cyber-physical systems (CPS), emerging as the most promising approach for extensive computing, processing, and remote control to physical entities, rely on reliable and heterogeneous (real-time and non-real-time) data exchanges among machines. However, for CPS exploiting shared network infrastructures, communication links may suffer from various vulnerabilities to harm real-time data exchanges. Providing robustness against link vulnerabilities consequently becomes the most critical requirement. By installing software defined networking (SDN) functions in existing network infrastructures, multiple communication paths can be practically formed to provide robustness if replicates of real-time data are simultaneously forwarded via multiple paths. Furthermore, if each non-real-time data packet is forwarded via distinct paths, packet congestion in certain paths can be avoided. However, the key to practice such SDN robust communications relies on an efficient resource utilization both in the time and the spatial domains. To further support high mobile machines, the efficient resource utilization shall be achieved by a low complexity scheme to rapidly adapt to varying environments. In this paper, we shall consequently develop mathematical foundations of a resource-optimal design for CPS. Achieving the minimum resource usage to support real-time data exchanges, our design further avoids the worst case packet congestion to trace the performance of non-real-time data exchanges to the optimum. By providing the most efficient and low complexity resource management, our design successfully practice robust big data exchanges in CPS.