1. Information about the paper


Pathak, Abhinav, Y. Charlie Hu, Ming Zhang, Paramvir Bahl, and Yi-Min Wang. “Fine-grained power modeling for smartphones using system call tracing.” InProceedings of the sixth conference on Computer systems, pp. 153-168. ACM, 2011.


Accurate, fine-grained online energy estimation and accounting of mobile devices such as smartphones is of critical importance to understanding and debugging the energy consumption of mobile applications. We observe that stateof-the-art, utilization-based power modeling correlates the (actual) utilization of a hardware component with its power state, and hence is insufficient in capturing several power behavior not directly related to the component utilization in modern smartphones. Such behavior arise due to various low level power optimizations programmed in the device drivers. We propose a new, system-call-based power modeling approach which gracefully encompasses both utilization-based and non-utilization-based power behavior. We present the detailed design of such a power modeling scheme and its implementation on Android and Windows Mobile. Our experimental results using a diverse set of applications confirm that the new model significantly improves the fine-grained as well as whole-application energy consumption accuracy. We further demonstrate fine-grained energy accounting enabled by such a fined-grained power model, via a manually implemented eprof, the energy counterpart of the classic gprof tool, for profiling application energy drain.


2. My review of the paper

  1. I agree that the authors of this paper have done a good work that is scalable.
  2. About flashing the kernel, I would say that to benefit most public users, the smartphone/mobile OS vendor might likely want to incorporate better power management on their future releases. They will more likely to do it if it has high chance of better power management. As they release eprof to public, someone might want to provide easy way to gain from this in form of easier flashing method.
  3. The real gain that we want is better power management that uses smallest power as possible for certain task. This power consumption estimation model is just a part way of gaining this. To really design such device, we really need to know how the hardware and software works. That’s why Apple leads on this one. Because Apple develop both hardware and software internally to tightly integrate the hardware and software. When I use Mac OS X on MacBook Pro retina, it can last for 9 hours. However, when I use latest Ubuntu 14.10 on MacBook Pro retina (same hardware), it only last for 3-4 hours. That’s why Microsoft bought Nokia and started their own Lumia to follow Apple’s development model. Google might (should) have done the same thing with Android flagship, which is Nexus devices.
  4. As I stated in point 3., then advocating open source device driver might really help.