Premium
On‐street smart parking networks at a fraction of their cost: performance analysis of a sampling approach
Author(s) -
Evenepoel S.,
Van Ooteghem J.,
Verbrugge S.,
Colle D.,
Pickavet M.
Publication year - 2014
Publication title -
transactions on emerging telecommunications technologies
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.366
H-Index - 47
ISSN - 2161-3915
DOI - 10.1002/ett.2776
Subject(s) - parking guidance and information , software deployment , computer science , parking lot , extrapolation , probabilistic logic , real time computing , wireless sensor network , transport engineering , engineering , computer network , civil engineering , mathematical analysis , mathematics , artificial intelligence , operating system
ABSTRACT Parking spot monitoring can be used for various smart parking applications. One particular use is detecting the city wide occupancy level of the parking spaces and redirecting cars accordingly to reduce congestion. The practice of monitoring parking spots is already widespread in modern off‐street parking lots. However, a large fraction of a city's parking capacity still stems from on‐street parking spots. These aren't yet monitored as commonly. This can be explained by higher deployment costs for on‐street implementations. In light of improving the adoption of on‐street parking spot monitoring, this paper evaluates a novel approach to reduce these deployment costs. The idea is to provide sensor readings on only a fraction of the parking spots and use extrapolation to calculate city wide saturation levels. This reduces the number of sensors that has to be installed and consequently lowers the network cost. A probabilistic model and corresponding performance measure were built to quantify the reliability of a sensor network that implements this approach. Dependent on the number of sensors deployed, the estimated saturation levels might be subject to extrapolation errors. An analysis of several deployments scenarios with decreasing numbers of sensors shows that significant cost reductions (up to a factor 30) are obtainable whilst retaining sufficient performance. These results exceed cost reductions obtained by alternative approaches where a switch is made from roadside to vehicle mounted sensors for monitoring parking spots. An additional comparative advantage of our approach is the reduced dependency on vehicles equipped with vehicle‐to‐vehicle or vehicle‐to‐environment communication units. Copyright © 2013 John Wiley & Sons, Ltd.