Context Sensing for Ubiquitous Computing

The Hong Kong University of Science and Technology
Department of Computer Science and Engineering


PhD Thesis Defence


Title: "Context Sensing for Ubiquitous Computing"

By

Mr. Wei SUN


Abstract

Ubiquitous computing is leading the third era of computing, after the ones 
represented by mainframe computers and personal computers. For many 
ubiquitous computing applications, context awareness is critical for 
performance optimization and user experience enrichment. Thanks to various 
tiny sensors embedded in devices such as smartphones, context information 
can be obtained by directly sensing the running environment. However, it 
is difficult to sense some certain types of context. In this thesis, we 
address three challenging topics of context sensing for ubiquitous 
computing, as briefed in the following.

Our first work focuses on knowing user's location in indoor environments. 
Most of the current solutions rely on Received Signal Strength (RSS) of 
wireless signals as location fingerprint, where fingerprint uniqueness 
with respect to locations is a basic assumption. However, due to practical 
limitations in real-world deployment, such assumption does not always 
hold, which we refer to as fingerprint ambiguity. In this work, we study 
the unexploited potential of user motion to resolve fingerprint ambiguity. 
Our basic idea is that user motion patterns collected by built-in sensors 
of mobile phones add to the fingerprint diversity. On this basis, we 
propose MoLoc, a motion-assisted localization scheme on mobile phones. The 
experimental results show that MoLoc achieves a significant improvement 
over fingerprinting-based methods.

Our second work addresses the problem of finding neighboring sensor nodes 
in the most energy-efficient way. We propose Hello, a generic flexible 
protocol for neighbor discovery. With an unrestricted parameter, it serves 
as a generic framework that incorporates existing deterministic protocols. 
Under the framework, we expose optimal parameters for either symmetric or 
asymmetric duty cycles, which is the first to our knowledge. The results 
from the simulation and real-world experiments show that Hello is highly 
energy-efficient under both symmetric and asymmetric duty cycles.

The last work attempts to detect sleep stages for users to understand 
their sleep quality. We present SleepHunter, a mobile service that 
provides a fine-grained detection of sleep stage transition for sleep 
quality monitoring and intelligent wake-up call. The rationale is that 
each sleep stage may be accompanied by specific and distinguishable body 
movements and acoustic signals. Leveraging the built-in sensors on 
smartphones, SleepHunter integrates these physical activities with sleep 
environment, inherent temporal relation and personal factors by a 
statistical model. Experimental results from over 30 sets of nocturnal 
sleep data show that our system provides a significant improvement in 
terms of detection accuracy when compared with existing actigraphy-based 
systems.


Date:			Tuesday, 19 May 2015

Time:			2:00pm - 4:00pm

Venue:			Room 3598
 			Lifts 27/28

Chairman:		Prof.  Ning Cai (IELM)

Committee Members:	Prof. Bo Li (Supervisor)
 			Prof. Kai Chen
 			Prof. Raymond Wong
 			Prof. Chin-Tau Lea (ECE)
 			Prof. Jianping Wang (CityU)


**** ALL are Welcome ****