Return to search

Distributed Functionality Allocation to Multi-core Computational Nodes

In previous years, there has been an increase in demand for new functionalities in the automotive and machinery domains. To meet these requirements, deploying more powerful single-core Electronic Control Units (ECUs) became necessary. However, single-cores may soon reach their limit in satisfying current performance and functionality needs. A promising solution lies in the use of multi-core ECUs. Although multi-cores are widely used in consumer electronics, their application in real-time systems has not yet been fully realized. This challenge stems from the fact that real-time systems impose strict timing requirements, making the transition to using multi-cores in real-time applications more challenging. This thesis investigates and analyzes methodologies to ensure that multi-core systems meet their real-time requirements. An extensive review of the state of the art is performed to determine different analysis approaches. Moreover, the thesis revises existing analyses, proposing refinements to enhance their correctness. Furthermore, the analyses are expanded to preemptive scheduling, as the majority of related work considers non-preemptive scheduling for multi-cores. Another scope of this thesis considers core interconnects. With that said, this thesis investigates two types of interconnects: bus and crossbar. An extensive analysis with evaluation is provided to compare the two. Furthermore, in order to investigate arbitration among cores for shared resources, two interconnect arbitration policies are used: First Come First Served and Round Robin. Theoretical considerations are laid out, and the evaluation of proposed approaches is presented. The proposed theories are tested against a typical automotive benchmarks and an industrial use case provided by Volvo Construction Equipment. The work culminates in assembling the analyses in a software tool developed in the C programming language. The tool is capable of running the aforementioned analyses and providing output based on the feasibility of the task set provided as input. Finally, the thesis discusses possibilities for refining the performed work and provides directions for future work.

Identiferoai:union.ndltd.org:UPSALLA1/oai:DiVA.org:mdh-66966
Date January 2024
CreatorsHrustic, Enesa, Kreho, Dzenan
PublisherMälardalens universitet, Inbyggda system
Source SetsDiVA Archive at Upsalla University
LanguageEnglish
Detected LanguageEnglish
TypeStudent thesis, info:eu-repo/semantics/bachelorThesis, text
Formatapplication/pdf
Rightsinfo:eu-repo/semantics/openAccess

Page generated in 0.0016 seconds