Energy Efficiency

Energy Efficiency has become more important in recent years. This is shown by the development of engery consumption of information and communication technology (ICT). In 2007, 10 % of the german energy consumption was generated by ICT. This trend is predicted to rise. Futhermore battery development cannot keep up with the ubiquitous and powerful mobile devices.

Research on hardware and low level software optimizations has been comprehensively explored. But the research on optimizing energy consumption on application level is still in its infancy. So in this project should be improved Energy Efficiency of applications by using reengineering services, like static and dynamic program analysis, and systematic code transformations.

Research Questions

Energy-Efficient Applications" is embedded to the University of Oldenburg, Department of Computing Sciences, research foci "Information Technology and Energy Efficiency" and "ExploIT Dynamics". Whereas "Information Technology and Energy Efficiency" defines the research domain, scientific approach and technologies are determined by "ExploIT Dynamics". The research questions in this project are:

• Energy consumption: Empowering applications to pass on energy-related information to operating system for smarter power management. Also identifying opportunities in software for better hardware utilization. Also one can take user behavior into account for choosing an energy-optimal strategy and identify futher opportunities by measuring and analyzing.
• Energy-consuming components: The energy consumption of single components has to be established, along with the degree of control the operating system or running applications have over their state. Then, applications would need to make information available to power management, regarding required components.
• Scheduling: With additional knowledge about the running applications, their current and future performance requirements, further optimization can be achieved.
• Energy-Awareness: With the information about the current energy-state of the hardware components, and the usage context (other running applications and the services they demand) applications could make purposeful decisions to conserve energy.

Reengineering Services

The described research questions could be solved with the following reengineering services:

• Energy consumption: To reduce energy consumption quantify, and certify energy efficiency of code should be detected. So code patterns and code smell detection for applying refactorings can be used to optimize the code. Also, the gather knowledge about running applications to make predictions regarding the system's behavior, helps to reduce the energy consumption.
• Energy-consumption components: An energy model can be established by an application at run-time with energy information of the different components. This can be coupled by dynamic analysis, e.g. recording execution traces to match usage scenarios to energy usage.
• Scheduling in mobile computing: Reengineering services can be used to collect these data, e.g. by monitoring running applications using code instrumentation. The resulting information is further analyzed for patterns, both of a single application, and between different applications running at the same time. This should avoid repeated wake ups from sleep states.
• Application usage analysis for energy-awareness: Application usage analysis aims at monitoring user activity and exploiting detected usage patterns for energy optimization. This make the application aware of energy status and usage.

Energy Management

To optimize the energy management of mobile devices, an energy abstraction layer can be embedded to the architectur. This layer is an abstract, platform-independent specification for measurement and management. Here, the application's and system's behavior are measured. Afterwards, the application's and system's behaivior are managed. Also a basic Energy API for aggregated functionality is provided.