University of Twente Student Theses

Login

Performance of real-time scheduling on sensor nodes : comparing scheduling algorithms, resource policies and energy conservation methods in AmbientRT

Bijlsma, T. (2006) Performance of real-time scheduling on sensor nodes : comparing scheduling algorithms, resource policies and energy conservation methods in AmbientRT.

[img]
Preview
PDF
1MB
Abstract:Wireless sensor networks are used to monitor the environment. In the new era of ubiquitous computing such networks are employed at different locations, to provide functionality and support to the users. These networks are composed out of small sensor nodes, which contain a fast microcontroller and sensors and are typically fed by a battery. To achieve predictable behavior the sensor nodes are equipped with a real-time operating system and a resource policy. A resource is a functionality that is possibly available to the tasks on the sensor node, for example the radio, a sensor or specific data. Furthermore energy conserving policies can be applied to utilize the idle time of the sensor node, to extend its life time. This report compares the performance of three real-time scheduling algorithms on a sensor node, namely the Earliest Deadline First (EDF), the Deadline Monotonic (DM) and the Rate Monotonic (RM) algorithms. The report also compares the transaction and the Nested Critical Section (NCS) resource policies, that can extend the examined scheduling protocols to manage shared resources. Two new policies to conserve energy in combination with real-time scheduling are introduced and examined. The Earliest Deadline First with Inheritance and Scaling (EDFIS) policy lowers the frequency of the microcontroller, while the Temporal Shutdown Scheduling (TSS) policy disables it when there is idle time. Performed tests show that the behavior of the EDF algorithm is most desirable. The transaction resource policy is the best solution when there are a lot of shared resources in the system. When there is an average amount of resources in the system, the NCS resource policy performs best. When a task set has a low utilization, most energy is conserved with the TSS policy. Above an average utilization the EDFIS policy shows the best power savings.
Item Type:Essay (Master)
Faculty:EEMCS: Electrical Engineering, Mathematics and Computer Science
Subject:53 electrotechnology
Programme:Electrical Engineering MSc (60353)
Link to this item:http://purl.utwente.nl/essays/57209
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page