Categorizing configuration parameters of smartphones for energy performance testing

Abstract

Energy performance testing in smartphones is a challenging task and the extent of exhaustive testing depends on the system configurations for different parameters and applications. In this paper, we propose a technique to classify the configuration parameters of a smartphone by partitioning them into two groups based on their maximum differential power (impact on power consumption). We validate the technique by applying it to four different smartphones: BlackBerry Bold 9700, BlackBerry Z10, Apple iPhone 3GS and Samsung Galaxy Nexus. The four devices represent a wide spectrum of devices with four operating systems (BB7, BB10, iOS, and Android), three makers (BlackBerry, Apple and Samsung), four hardware platforms, and relatively old and new devices.