This book introduces the concepts of mobility data and data-driven urban
traffic monitoring. A typical framework of mobility data-based urban
traffic monitoring is also presented, and it describes the processes of
mobility data collection, data processing, traffic modelling, and some
practical issues of applying the models for urban traffic monitoring.
This book presents three novel mobility data-driven urban traffic
monitoring approaches. First, to attack the challenge of mobility data
sparsity, the authors propose a compressive sensing-based urban traffic
monitoring approach. This solution mines the traffic correlation at the
road network scale and exploits the compressive sensing theory to
recover traffic conditions of the whole road network from sparse traffic
samplings. Second, the authors have compared the traffic estimation
performances between linear and nonlinear traffic correlation models and
proposed a dynamical non-linear traffic correlation modelling-based
urban traffic monitoring approach. To address the challenge of involved
huge computation overheads, the approach adapts the traffic modelling
and estimations tasks to Apache Spark, a popular parallel computing
framework. Third, in addition to mobility data collected by the public
transit systems, the authors present a crowdsensing-based urban traffic
monitoring approach. The proposal exploits the lightweight mobility data
collected from participatory bus riders to recover traffic statuses
through careful data processing and analysis. Last but not the least,
the book points out some future research directions, which can further
improve the accuracy and efficiency of mobility data-driven urban
traffic monitoring at large scale.
This book targets researchers, computer scientists, and engineers, who
are interested in the research areas of intelligent transportation
systems (ITS), urban computing, big data analytic, and Internet of
Things (IoT). Advanced level students studying these topics benefit from
this book as well.
