Road traffic congestions result in important economical and
productivity losses, as well as an increasing environmental
impact. To reduce their negative effects, novel Intelligent
Transportation Systems (ITS) are currently being investigated.
One of the critical aspects that will enable the design of
efficient traffic management policies will be the capacity to
effectively monitor the traffic conditions and rapidly detect
traffic congestion. Conventional traffic monitoring solutions
based on infrastructure sensors include inductance loops, video
and image processing, and microwave radars. These types of
detectors provide fixed-point or short-section traffic
information that is extracted from vehicles passing the
detection zone. They generally report data about the vehicles’
volume and lane occupancy, that combined, can be used to
estimate the vehicles’ speed. One of the main limitations of
point detection technologies is that the traffic estimates are
based on measurements taken at a specific location that might
not provide an accurate representation of the traffic conditions
over larger road segments. The design of point-based traffic
monitoring solutions is therefore characterized by a trade-off
between the traffic estimates accuracy and the number of
deployed infrastructure sensors.