Key points are not available for this paper at this time.
Recently, large scale cyber physical systems (LS-CPS) leverage network-cores provided by application providers (APs) to carry out analytics. These CPS-APs uses the automated cloud to gather traffic data streams, thereby reducing infrastructure and maintenance costs. However, network reliability and maintainability considering the arrival rate of user application requests presents a major challenge such as Edge-to-Fog and Fog-to-Cloud resource auto-scaling constraints. QoS dynamic reliability, as well as its flexible management, could offer an efficient network for CPS. In this paper, CloudMesh CPS architecture is presented as a promising solution to support services like Input-Output (IO) data stream, traffic engineering, service function optimization and software defined network monitoring in CPS IPv6 data-center core. Dynamic reliability modeling for QoS parameter scaling based on observable history is presented. Hence, QoS proactive auto-scaling algorithm (PASCQA) embedded with a heuristic predictor is introduced in CloudMesh CPS. IO streams are captured by the predictor which analyzes the QoS history for reliable global performance. Finally, the work realized the critical performance aspects of CPS architecture and analyzed the effects of QoS parameter configuration for real-time deployment. The implementation results show that CloudMesh offers satisfactory QoS metrics compared with Fat-tree-1 and Fat-tree-2 for CPS applications.
Kennedy Chinedu Okafor (2019) studied this question.