Diagnosis of Smartphone Battery and Charger

Smartphone users frequently encounter battery and energy problems. From the popular Internet blogs, we have identified two issues that are increasingly being reported by the users; sudden drop in the battery level and disgraceful shutdown of the device even with high battery levels being reported to the user (even at 80%) while discharging. These observations are reported across different smartphone models, and even for laptops. This disgraceful shutdown may bar users from their scheduled phone activities and result in data loss. From a user’s perspective, the remaining battery life of a smartphone may even converge to the monetary value. Many smartphone manufacturers have recently intro- duced battery replacement programs that cover batteries that have a reduced capacity, typically below 80%. The current smartphone battery discussion pertains to the following questions: Why does the battery level fluctuate? Is the battery faulty? Is the problem due to an operating system upgrade or installing/upgrading an application?.

The answers to the earlier questions are related to the full charge capacity (FCC) of smartphone battery, as we demonstrated in prior work. FCC is the maximum amount of charge that an empty battery can hold. As the battery of a smartphone ages, the full charge capacity decreases with the utilization. FCC is typically modeled as a function of the number of charging cycles or measured with Coulomb counter. This functionality or measurement capability resides inside a smart battery. The battery shares this estimate as percentage with the hosting device, such as smartphone (see Section II). Therefore, an indication of capacity loss along with the battery level would allow more accurate state of charge (SOC) estimation, which improves the reliability of battery-powered smartphones. Different applications may use this information in different ways. For instance, a cheaper and simple voltage-based fuel gauge can estimate FCC while charging and apply it for estimating SOC instead of applying complex charging cycle-based learning. Similarly, a simple Coulomb counter-based fuel gauge can calibrate FCC even when the battery is partially charged. The other applications are power consumption modeling, sophisticated energy-aware resource scheduling mechanisms by the system and different applications. 

We investigate the behavior of smartphone batteries while charging and demonstrate that battery voltage and charging rate information can together characterize the FCC of a battery. We propose a new method for accurately estimating FCC without exposing low-level system details or introducing new hardware or system modules. We further propose and implement a collaborative FCC estimation technique that builds on crowd-sourced battery data. The method finds the reference voltage curve and charging rate of a particular smartphone model from the data and then compares with those of an individual device. After analyzing a large data set towards a crowd-sourced rate vs. FCC model, we report that 55% of all devices and at least one device in 330 out of 357 unique device models lost some of their FCC. For some old device models, the median capacity loss exceeded 20%. 

The models further enable debugging the performance of smartphone charging. We implement an algorithm, called BatterySense, which utilizes crowd-sourced rate to detect abnormal charging performance, estimate FCC of the device battery, and detect battery changes.