T-L Plane Abstraction-Based Energy-Efficient Real-Time Scheduling for Multi-Core Wireless Sensors

Area of Science:

• Computer Science
• Electrical Engineering
• Embedded Systems

Background:

• Energy efficiency is crucial for wireless sensor networks (WSNs).
• Multi-core processors in WSN nodes necessitate energy-efficient real-time scheduling algorithms.
• T-L plane-based schemes are optimal for periodic real-time tasks on multi-cores but lack energy-saving extensions.

Purpose of the Study:

• To propose a novel T-L plane-based algorithm for energy-efficient real-time scheduling on multi-core sensor nodes.
• To integrate dynamic power management (DPM) into T-L plane scheduling.
• To address inherent limitations of existing T-L plane algorithms, such as processor mode transition overhead and idle time fragmentation.

Main Methods:

• Development of a new T-L plane-based scheduling algorithm incorporating dynamic power management.
• Focus on minimizing processor mode transition overhead.
• Techniques to reduce idle time fragmentation in T-L plane scheduling.

Main Results:

• The proposed algorithm demonstrates effectiveness in energy-efficient real-time scheduling.
• It successfully reduces processor mode transition overhead.
• It mitigates idle time fragmentation compared to other T-L plane methods.
• Experimental results validate the algorithm's performance against existing energy-aware scheduling methods.

Conclusions:

• The novel T-L plane-based algorithm enhances energy efficiency in multi-core WSNs.
• The approach effectively manages dynamic power management for real-time tasks.
• It offers a significant improvement over existing energy-aware scheduling techniques within the T-L plane framework.