Paper # 4

Energy-aware adaptation for mobile applications

Problems

The paper discusses how a collaborative relationship between the operating
system and applications can be used to conserve energy and used to meet
user-specified goals for battery duration.

Strengths and New Ideas

1. Fidelity -- fidelity, as used in this paper, is described as the degree
to which data presented at a client matches the reference copy at a server.
Reduction in fidelity for different data types accross different applications
results in conservation of energy. This means that applications that can
adapt and change the fidelity can use less energy.

2. The paper is extremely easy to understand and uses a simple and intuitive idea,
that of fidelity, to help in energy conservation.

3. The authors choose the adaptive test applications carefully--a video viewer, a
speech recognizer, a map viewer, and a browser. These are all applications
that are used extensively on mobile computing devices and therefore most applicable
to do energy conservation tests on (for mobile devices).

4. The role of Odyssey and how it interacts with each of these applications is 
clearly stated although at a high level. This helps in understanding how they 
implement their origninal idea of collaboration between the Operating System
and applications.

5. Each of the experiments are designed and executed carefully to give a comprehensive picture
of the amount of energy savings that are possible through hardware as well as
different measures of fidelity.

6. The authors point out that there is siginificant variance in effectiveness of
fidelity reduction across data objects which implies that there is a chance
that data object dependent fidelity management could be introduced.

7. Fidelity effectiveness varies across applications--this implies that it might
be useful to concentrate adaptive application development in only those areas
that offer the highest levels of energy conservation through fidelity control.

8. The fact that user defined goals were met through the use of Odyssey and
adaptive applications is promising and could be highly beneficial to the
mobile computing industry. This also has  tremendous commercial potential.

Weaknesses

1. There is no mention of the cost of complexity associated with the development
of Odyssey like systems as well as adaptive applications. Is it fairly easy to
develop such applications? If not, then what are the issues? This should have been
discussed.

2. The authors did not discuss how they came up with their choice of how to control
fidelity for each application. Since effectiveness is dependent on this choice, it is
important to discuss this aspect.

3. While the applications chosen for the experiments were different, they were small
in number (only 4). More applications should be tested and the results analyzed.