TY - GEN
T1 - Distributed or centralized traffic advisory systems-the application's take
AU - Otto, John S.
AU - Bustamante, Fabian E
PY - 2009
Y1 - 2009
N2 - We consider the problem of data dissemination in vehicular networks. Our main goal is to compare the applicationlevel performance of fully distributed and centralized data dissemination approaches in the context of traffic advisory systems. Vehicular networks are emerging as a new distributed system environment with myriad promising applications. Wirelesslyconnected, GPS-equipped vehicles can be used, for instance, as probes for traffic advisory or pavement condition information services with significant improvements in cost, coverage and accuracy. There is an ongoing discussion on the pros and cons of alternative approaches to data distribution for these applications. Proposed centralized, or infrastructure-based, models rely on road-side equipment to upload information to a central location for later use. Distributed approaches take advantage of the direct exchanges between participating vehicles to achieve higher scalability at the potential cost of data consistency. While distributed solutions can significantly reduce infrastructures' deployment and maintenance costs, it is unclear what the impact of "imprecise" information is to an application or what level of adoption is needed for this model to be effective. This paper investigates the inherent trade-offs in the adoption of distributed or centralized approaches to a traffic advisory service, a commonly proposed application. We based our analysis on a measurements study of signal propagation in urban settings and an extensive simulation-based experimentation in the Chicago road network.
AB - We consider the problem of data dissemination in vehicular networks. Our main goal is to compare the applicationlevel performance of fully distributed and centralized data dissemination approaches in the context of traffic advisory systems. Vehicular networks are emerging as a new distributed system environment with myriad promising applications. Wirelesslyconnected, GPS-equipped vehicles can be used, for instance, as probes for traffic advisory or pavement condition information services with significant improvements in cost, coverage and accuracy. There is an ongoing discussion on the pros and cons of alternative approaches to data distribution for these applications. Proposed centralized, or infrastructure-based, models rely on road-side equipment to upload information to a central location for later use. Distributed approaches take advantage of the direct exchanges between participating vehicles to achieve higher scalability at the potential cost of data consistency. While distributed solutions can significantly reduce infrastructures' deployment and maintenance costs, it is unclear what the impact of "imprecise" information is to an application or what level of adoption is needed for this model to be effective. This paper investigates the inherent trade-offs in the adoption of distributed or centralized approaches to a traffic advisory service, a commonly proposed application. We based our analysis on a measurements study of signal propagation in urban settings and an extensive simulation-based experimentation in the Chicago road network.
UR - http://www.scopus.com/inward/record.url?scp=70449574091&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=70449574091&partnerID=8YFLogxK
U2 - 10.1109/SAHCN.2009.5168927
DO - 10.1109/SAHCN.2009.5168927
M3 - Conference contribution
AN - SCOPUS:70449574091
SN - 9781424429080
T3 - 2009 6th Annual IEEE Communications Society Conference on Sensor, Mesh and Ad Hoc Communications and Networks, SECON 2009
BT - 2009 6th Annual IEEE Communications Society Conference on Sensor, Mesh and Ad Hoc Communications and Networks, SECON 2009
T2 - 6th Annual IEEE Communications Society Conference on Sensor, Mesh and Ad Hoc Communications and Networks, SECON 2009
Y2 - 22 June 2009 through 26 June 2009
ER -