IEEE Communications Magazine - June 2017 - page 148

IEEE Communications Magazine • June 2017
146
0163-6804/17/$25.00 © 2017 IEEE
A
bstract
Fog-based vehicular crowdsensing is an emerg-
ing paradigm where vehicles use onboard sensors
to collect and share data with the aim of measuring
phenomena of common interest. Unlike traditional
mobile crowdsensing, fog nodes are introduced
specifically to meet the requirements for loca-
tion-specific applications and location-aware data
management in vehicular ad hoc networks. In this
article, we examine the architecture, applications,
and especially security, privacy, and fairness of
fog-based vehicular crowdsensing. Specifically, we
first introduce the overall infrastructure and some
promising applications, including parking naviga-
tion, road surface monitoring, and traffic collision
reconstruction. We then study the security, privacy,
and fairness requirements in fog-based vehicular
crowdsensing, and describe the possible solutions
to achieve security assurance, privacy preserva-
tion, and incentive fairness. By defining interesting
future directions, this article is expected to draw
more attention into this emerging area.
I
ntroduction
The integration of sensors and embedded com-
puting devices triggers the emergence of mobile
crowdsensing services [1], which allows individ-
uals to cooperatively collect and share data and
extract information to measure and map phe-
nomena of common interest using sensing and
communication technologies. With the increasing
popularity of mobile devices, mobile crowdsens-
ing becomes a broad range of sensing paradigms
nowadays. For example, an iPhone 6S can sense
the environment with a rich set of sensors, includ-
ing a camera, GPS, a proximity sensor, and a
barometric sensor, to generate and share the
sensing reports with interested parties [2].
Similar to mobile phones, modern vehicles are
also equipped with onboard sensors and wire-
less communication devices [3], such as camer-
as, GPS, tachographs, lateral acceleration sensors,
and onboard units (OBUs), providing fundamental
capability and feasibility of vehicular crowdsensing.
By using OBUs and sensing devices, vehicles can
not only periodically report the driving information
(e.g., location, real-time speed, and driving video)
but also incidentally provide traffic conditions, road
conditions, and weather conditions for transpor-
tation planning, road system design, traffic signal
control, and so on [4]. The approach of raw data
acquisition through vehicular crowdsensing signifi-
cantly reduces the financial and time cost for data
customers. With the development of electric devic-
es in vehicles, the sensing data become increas-
ingly fine-grained and complex, so the data from
vehicles are extended to support more applica-
tions, such as vehicle fault diagnostic, vehicle noise
pollution detection, and air quality forecast. Mean-
while, fine-grained data collection increases the
burden on data transmission and centralized data
management. The cloud server has to maintain
and process data for supporting vehicular crowd-
sensing services. Nevertheless, local relevance is
one of the important features of vehicle-generated
data, which means that the sensing data have their
own spatial scope and explicit lifetime of utility.
For example, traffic congestion information may
only be valid for 30 minutes and of interest to the
vehicles that are approaching a traffic jam area.
Vehicle-generated contents are also local interests,
indicating that the traffic and road condition infor-
mation of a specific region are only of interest to
the vehicles in or near that region. Therefore, cen-
tralized data management is not recommended,
and the sensing data should be classified according
to the spatial-temporal information.
Fog computing is a particularly attractive para-
digm [5] that utilizes network edge devices to carry
out a substantial amount of storage, communica-
tion, and computing close to the mobile devices,
so it is not necessary to send all data all the way to
the cloud. With temporary data storage, computing,
and processing, the constraints of the information
interactions between the cyber world and physi-
cal world, in terms of latency, load balancing, and
fault tolerance, can be released. These appealing
advantages trigger the emergence of fog-based
vehicular crowdsensing (FVCS). On behalf of local
servers, vehicular fog nodes can temporarily store
and analyze the sensing data uploaded by vehicles
to provide local services (e.g., real-time navigation,
parking space reservation, and restaurant recom-
mendation). They can process the data locally and
pass the results to interested vehicles quickly, there-
by saving unnecessary wireless bandwidth for trans-
mitting the raw data to a remote cloud server and
also supporting location-aware data management.
Therefore, FVCS not only inherits the advantages
of mobile crowdsensing [1], but also integrates fog
computing to have unique characteristics, including
location awareness, geo-distribution, and commu-
nication efficiency. However, security and privacy
Security, Privacy, and Fairness in
Fog-Based Vehicular Crowdsensing
Jianbing Ni, Aiqing Zhang, Xiaodong Lin, and Xuemin (Sherman) Shen
S
ustainable
I
ncentive
M
echanisms
for
M
obile
C
rowdsensing
The authors examine the
architecture, applications,
and especially security,
privacy, and fairness
of fog-based vehicular
crowdsensing. They
introduce the overall
infrastructure and some
promising applications,
including parking naviga-
tion, road surface moni-
toring, and traffic collision
reconstruction, then study
the security, privacy, and
fairness requirements,
and describe the possi-
ble solutions to achieve
security assurance, privacy
preservation, and incen-
tive fairness.
Jianbing Ni is with the University of Waterloo; Aiqing Zhang is with Anhui Normal University and the University of Ontario Institute of Technology;
Xiaodong Lin is with the University of Ontario Institute of Technology; Xuemin (Sherman) Shen is with the University of Waterloo.
Digital Object Identifier:
10.1109/MCOM.2017.1600679
1...,138,139,140,141,142,143,144,145,146,147 149,150,151,152,153,154,155,156,157,158,...228
Powered by FlippingBook