Energy-Aware Scheduling in Multiprocessor Real-Time Systems
| Author | Awwad, Noor Adnan | |
|---|---|---|
| metadata.dc.contributor.department | UNIVERSITY OF BAHRAIN College of Information Technology | |
| metadata.dc.date.accessioned | 2026-09-20T15:48:37Z | |
| Publication date | 2016 | |
| Degree level | Master | |
| Abstract | Abstract : Energy consumption is becoming one of the most critical issues in various system designs. In multiprocessor systems, the energy consumption is highly affected by task allocation approaches. In the last decades designing an efficient system with lower power consumption has captured the attention of the research community in the area of computer science. The problem is considered as a design goal in limited power systems and mobile systems, particularly for real-time systems. Power-aware scheduling of real-time tasks using dynamic scaling on multi-processor systems is addressed in this thesis. Real-time task allocation problem is presented as an optimization problem subjected to a number of constraints. Many algorithms have been developed to solve the optimization problem in order to achieve optimal or semi optimal results. Nature-inspired algorithms such as Genetic algorithm (GA) and Cuckoo search (CS) are adapted in this research to solve the underlying problem. Parameter sensitivity analysis is used to decide the most appropriate parameters' value and combinations for both algorithms. Several variations of the crossover and mutation operators in GA are proposed. For the CS algorithm, an approach for determining suitable values of its parameters is suggested. The research parameters' value and combinations for both algorithms. Several variations of the crossover and mutation operators in GA are proposed. For the CS algorithm, an approach for determining suitable values of its parameters is suggested. The research emphasizes the effect of different parameters and operators on algorithms performance. The efficiency of algorithms is compared in terms of solution quality and execution time. The most appropriate parameters' value and combinations are then determined for each algorithm. A variation of basic CS (called CS with penalty) is also proposed in which infeasible solutions are allowed to be part of the population with a penalty. The adapted approaches provide savings up to 89% for the best cases in power consumption. GA and CS in most cases produced similar results in terms of solution quality. However GA slightly outperforms CS in terms of execution time. In general, simulation results on multiprocessor system composed of 5 processors show that GA performs better than CS by a percentage up to 58% also GA is performing better than CS with penalty by percentage up to 50% in almost all cases. | |
| metadata.dc.identifier.other | Drupal node 3748 | |
| metadata.dc.identifier.other | https://digitalrepository.uob.edu.bh/en/dar/energy-aware-scheduling-multiprocessor-real-time-systems | |
| metadata.dc.identifier.uri | http://165.227.56.122/handle/123456789/6776 | |
| Language | en | |
| Link | https://digitalrepository.uob.edu.bh/en/_flysystem/fedora/2025-07/energy-aware_scheduling_in_multiprocessor_real-time_systems_1.pdf | |
| Title | Energy-Aware Scheduling in Multiprocessor Real-Time Systems | |
| Type | Master's Thesis |
Files
Original bundle
1 - 1 of 1
Loading...
- Name:
- node-3748-1-energy-aware_scheduling_in_multiprocessor_real-time_systems_1.pdf
- Size:
- 17.35 MB
- Format:
- Adobe Portable Document Format