The increasing complexity and adaptive dynamic behavior of cyber-physical systems, such as advanced driver assistance systems (ADAS) or service robotics, require novel embedded hardware/software solutions. Especially, the dynamic behavior at runtime needs an approach providing adaptation to changing demands in terms of real-time requirements, data throughput, safety and security. One representative example can be found in robotic systems, where changing situations are handled with image processing algorithms for object detection and tracking. Here, the changing situations would recommend besides the change of the algorithm also a change in the hardware architecture, e.g. the adaptation of accelerators for specific algorithms.
Self-Adaptive Multiprocessor Systems-on-Chip for Cyber-Physical Systems
The complexity and the high demand for real-time and energy efficient computation of such adaptive dynamic systems can only be solved with parallel and runtime adaptive hardware/software solutions and optimized design- and programming tools. Traditional embedded multicore systems can only handle a task migration from core to core in order to balance the workload of an individual processor. However, only migrating software is not sufficient to find the optimal point of operation. Changes of the processor architecture and certainly the communication infrastructure between cores would be highly beneficial. This feature can be provided by heterogeneous self-adaptive multiprocessor systems-on-chip (MPSoCs), where each component can be adapted according to application demands.
This presentation shows concepts and realizations for such a modern approach including a novel computer architecture, a simulation environment, the design/programming methods and the runtime management. The importance of such a novel hardware/software solution is shown with several research projects in the automotive and robotics domain.