A Survey of Mobile Power Management

Posted by Kaiyuan Chen on May 2, 2018

The following blog tries to survey papers of Mobile Power Management.

Never mind, a killer survey: https://onlinelibrary.wiley.com/doi/full/10.1002/dac.3234.

Measurement:

Except HW Monsoon and SW power tutor:

battery historian 2.0(by google)

https://github.com/google/battery-historian

Estimation:

Application Processor(Multicore):online estimation of application-level power consumption. It has mean absolute percentage error (MAPE) of 5.1%

https://www.sciencedirect.com/science/article/pii/S1383762117301947

(2017)

Power Estimation:

https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=6629277

the difference between the measured data and the calculated power numbers are very small, with all differences less than 18%

(2013)

full-system simulator:

it provides detailed performance metrics for an entire system.

26.8% error rate for various network packet loss rates power modeling accuracy within 12.8% error rate

https://ieeexplore.ieee.org/document/7482099/

Optimization

stops RILD(LTE/GSM):https://ieeexplore.ieee.org/document/7980184/ can reduce 42%

(2017)

Others:

Context Aware app profiling

a self-adaptive framework that haves three power profiles that react to battery level, status and context

https://ieeexplore.ieee.org/document/6906452/

(2014)