DocumentCode :
656218
Title :
Extending Battery Life of a Multi-buffered, Single-Threaded Processor in a Mobile Computing Device
Author :
Khogali, Rashid ; Das, Olivia
Author_Institution :
Dept. of Electr. & Comput. Eng., Ryerson Univ., Toronto, ON, Canada
fYear :
2013
fDate :
1-4 Oct. 2013
Firstpage :
817
Lastpage :
825
Abstract :
We introduce an online speed-scaling algorithm that is used to determine the optimum processing rate of executing a set of N jobs by a single processor of a mobile computing device under the single-threading (multi-buffered) computing architecture. We consider heterogeneous tasks that could differ in computation volume, memory and processing requirements. By using speed-scaling, where the processor´s speed is able to dynamically change within hardware and software processing constraints, the algorithm explicitly determines the optimum processing rate of executing each task. This optimum processing rate was found to be a function of the number of ´alive´ tasks (N), the remaining battery energy percentage, the processor´s energy inefficiency coefficient, the unit price of response time and lastly, the unit price of energy. The algorithm allows the user or OS to specify the unit cost of energy and response time for executing all tasks. The algorithm has an operation mode where all tasks´ unit cost of energy is also heuristically affected by the device´ remaining battery energy percentage in accordance with the micro-economic laws of demand and supply. We synthesize the algorithm by analytically minimizing the total cost of both response time and energy consumption of tasks. We also consider other conventional performance metrics to evaluate the algorithm. Using numerical simulations, we show that when the remaining battery energy percentage is factored, the algorithm performs slightly slower (mildly more slower when the battery is almost drained out), but consumes far less energy, can complete significantly more jobs and ultimately allows the mobile computing device to last longer on the go.
Keywords :
microprocessor chips; mobile computing; power aware computing; battery energy percentage; computation volume; extending battery life; memory requirements; microeconomic laws; mobile computing device; multi-buffered; multibuffered computing architecture; numerical simulations; online speed scaling algorithm; optimum processing rate; processing requirements; processor energy inefficiency coefficient; single threaded processor; software processing constraints; Parallel processing; battery energy; local; mobile; single processor computing; single-threaded; speed scaling;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Parallel Processing (ICPP), 2013 42nd International Conference on
Conference_Location :
Lyon
ISSN :
0190-3918
Type :
conf
DOI :
10.1109/ICPP.2013.96
Filename :
6687421
Link To Document :
بازگشت