Abstract
In this paper an efficient algorithm is proposed which optimizes periodic message scheduling in a real-time multiprocessor system. The system is based on a many-core single-chip computer architecture and uses a multistage baseline network for inter-core communication. Due to its basic architecture, internal blockings can occur during data transfers, i.e. the baseline network is not real-time capable by itself. Therefore, we propose a scheduling algorithm that may be performed before the execution of an application in order to compute a non-blocking schedule of periodic message transfers. Additionally, we optimize the clock rate of the network subject to the constraint that all data transfers can be performed in a non-blocking way. Our solution algorithm is based on a generalized graph coloring model and a randomized greedy approach. The algorithm was tested on some realistic communication scenarios as they appear in modern electronic car units. Computational results show the effectiveness of the proposed algorithm.
| Original language | English |
|---|---|
| Pages (from-to) | 374-382 |
| Number of pages | 9 |
| Journal | Journal of Systems Architecture |
| Volume | 61 |
| Issue number | 8 |
| DOIs | |
| Publication status | Published - 14 Sept 2015 |
| Externally published | Yes |
Keywords
- Baseline network
- Graph coloring
- Message scheduling
- Periodic scheduling
- Real time
Fingerprint
Dive into the research topics of 'Message scheduling for real-time interprocessor communication'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver