Run-time resource management based on design space exploration

Abstract

A main challenge in today's embedded system design is to find the perfect balance between performance and power consumption. This paper presents a run-time resource management framework for embedded heterogeneous multi-core platforms. It allows dynamic adaptation to changing application context and transparent optimization of the platform resource usage following a distributed and hierarchical approach. A Global Resource Manager (GRM) is running in parallel with the central manager of the application on the host processor of the platform. Each IP core of the platform can execute its own Local Resource Manager (LRM), and the GRM conforms to practices of each LRM. The operating points managed by the GRM are identified in a design-space exploration phase as a set of Pareto-optimal configurations of the application and their impacts with regards to the quality of experience, performance and energy consumption. The GRM has already been integrated in a POSIX version of an audio-driven video surveillance application in order to maximize its QoE parameters with respect to the battery duration and the energy budget of the platform, used to analyze the GRM efficiency.