HETERO-SCHED: A Low-Overhead Heterogeneous Multi-core Scheduler for Real-Time Periodic Tasks

Abstract

Real-time systems are increasingly being implemented on heterogeneous multi-core platforms to efficiently cater to their diverse and high computation demands. However, devising efficient resource allocation strategies for real-time tasks on heterogeneous platforms has traditionally proved to be a challenging as well as a computationally expensive problem. As a consequence, today we face a severe dearth of low overhead real-time scheduling techniques which are applicable to heterogeneous platforms. Hence, this paper proposes an effective low-overhead heuristic approach called HETEROSCHED, for scheduling a set of periodic tasks executing on a heterogeneous multi-core system. The proposed approach first applies deadline partitioning to obtain a set of discrete time slices. Over each such time-slice, HETERO-SCHED conducts the following two phase operation: First, it determines the fractions of the computation demand of each task to be assigned onto the platform. Next, it assigns valid start and finish times to all tasks, according to the allocation prescribed in the first phase. Experimental studies show that our proposed scheduling mechanism is able to significantly improve acceptance ratios for task sets, compared to the state-of-the-art.