Energy-, Latency and Resilience-aware Networking

点击翻译按钮获取中文摘要

Explicitly increasing the predictability of the network behavior by bringing latency- and resilience-awareness to the entire cyber-physical network stack has been the major achievement of the project LARN and a significant contribution to the priority program "Cyber-Physical Networking".The continuation project will -- in collaboration within the priority program but also with a sister proposal under the US National Science Foundation's Cyber-Physical Systems program -- extend these results into two important directions:First, energy demand will be addressed. As well for the wired backbone (due to the sheer amount of data) as for wireless nodes (due to the restrictions of battery powered devices) energy-awareness and the optimization of the networks stack towards the reduction of its energy demand becomes inevitable. We will extend the awareness of the overall CPN to besides latency and resilience include energy.Second, we will exploit the predictability to optimally serve cyber-physical systems and the quality of control in those systems. Other than making control algorithms robust against random losses and random delays, we will enable statistical shaping of the network behavior to optimize the coaction of CPNs (networks) and CPSs (systems). For this we will exploit input from control systems researchers within the priority program but also with US colleagues from Wayne State University, Michigan, USA.Hence, we will continue to unite the disciplines "Operating Systems/Real-Time Processing" and "Telecommunications/Information Theory" and complement this through the inclusion of an explicit collaboration with "Control Systems". Two application scenarios (distributed, networked power microgrids and autonomous vehicles) will be addressed and demonstrated by means of a) the provision of RNAs (Reliable Network Atoms) and b) the validation in a real-world wide area network testbed. Both, the RNAs and the testbed will remain to be available to other partners of the SPP and to collaborating projects and researchers from the NSF's CPS program.