On Adaptive Resource Allocation for Complex Real-Time Applications

Resource allocation for high-performance real-time applications is challenging due to the applications’ datadependent nature, the dynamic changes in their external environment, and the limited resources available of the embedded systems on which they run. These challenges may be met by use of Adaptive Resource Allocation (ARA) mechanisms that can promptly adjust resource allocation to changes in applications’ resource needs, whenever there is a risk of failing to satisfy the application’s timing constraints. Although not decided by the application, such adjustments satisfy the application’s adaptation capabilities. ARA eliminates the need for ‘over-sizing’ real-time systems to meet worst-case application needs. This paper proposes an application model used to describe the application’s resource needs and its adaptation capabilities. The model also describes the runtime variation of application needs. The paper also proposes a satisfiability-driven set of performance metrics for capturing the impact of ARA mechanisms on the performance of real-time applications. The relevance of the proposed metrics set is demonstrated experimentally, using an adaptive, synthetic application designed to represent time-critical applications in C3I systems.