Resource-aware task allocation and scheduling for segbus platform

Abstract

In this work, we propose an integrated task allocation and scheduling mechanism to minimize the resource contention and the processing latency for application running on the SegBus platform. The transactions are classified as local and cross border SPLIT transactions. The hybrid scheduling approach implemented by hierarchal arbiter code structure shows significant improvement in system performance. The interrupt scheduling has been implemented to further enhance system performance. A H.264 video encoder application has been used to verify the proposed technique, showing a large improvement in system throughput.