Diversity in Smartphone Usage
11. CONCLUSIONS
By studying 255 users of two different smartphone platforms,
we comprehensively characterized user activities and
their impact on network and battery. We quantify many
hitherto unknown aspects of smartphone usage. User diversity
is an overarching theme in our findings. For instance,
different users interact with their phones 10-200 times a day
on average; the mean interaction length of different users
193
is 10-250 seconds; and users receive 1-1000 MB of data per
day, where 10-90% is received as part of active use.
This extent of user diversity implies that mechanisms that
work for the average case may be ineffective for a large fraction
of the users. Instead, learning and adapting to user
behaviors is likely to be more effective, as demonstrated by
our personalized energy drain predictor. We show that qualitative
similarities exist among users to facilitate the development
of such mechanisms. For instance, the longer the
user has not interacted with the phone, the less likely it is
for her to start interacting with it; and application popularity
for a user follows a simple exponential distribution.
Our study points to ways in which smartphone platforms
should be enhanced. Effective adaptation will require future
platforms to support light-weight tools that monitor and
learn user behaviors in situ. It will also require them to
expose appropriate knobs to control the behavior of lowerlevel
components (e.g., CPU or radio)