ABSTRACT
The feedback-control dynamic resource allocation heuristics impose some requirements on the target system, which should guarantee that the appropriate input data is available and that the generated output can be used to perform the proper resource allocation. This chapter presents the task mapping process which is comprised of the resource allocation and task scheduling. The resource allocation process is executed on a particular processing unit. The chapter proposes an algorithm to map firm real-time tasks into multi-core systems dynamically, using dynamic voltage and frequency scaling to decrease energy dissipation in cores. According to simulation results, the proposed method leads to more than 55% of dynamic energy reduction. The obtained results have confirmed the accumulating (integrating) nature of the process, and thus the accumulating process version of Approximate M-constraint Integral Gain Optimization tuning formulas have been applied to choose the proper values of the PID controller components.
