System adaptivity is becoming an important feature of modern embedded multiprocessor systems. To achieve the goal of system\r\nadaptivity when executing Polyhedral Process Networks (PPNs) on a generic tiled Network-on-Chip (NoC) MPSoC platform, we\r\npropose an approach to enable the run-time migration of processes among the available platform resources. In our approach,\r\nprocess migration is allowed by a middleware layer which comprises two main components. The first component concerns the\r\ninter-tile data communication between processes. We develop and evaluate a number of different communication approaches\r\nwhich implement the semantics of the PPN model of computation on a generic NoC platform. The presented communication\r\napproaches do not depend on the mapping of processes and have been implemented on a Network-on-Chip multiprocessor\r\nplatform prototyped on an FPGA. Their comparison in terms of the introduced overhead is presented in two case studies\r\nwith different communication characteristics. The second middleware component allows the actual run-time migration of PPN\r\nprocesses. To this end, we propose and evaluate a process migration mechanism which leverages the PPN model of computation to\r\nguarantee a predictable and efficient migration procedure. The efficiency and applicability of the proposed migration mechanism\r\nis shown in a real-life case study.
Loading....