On optimal service directory selection in urban vehicular networks

Authors:
Fenggang Wu;Hongzi Zhu;Jia-Liang Lu;Min-You Wu
Affiliations:
Shanghai Jiao Tong University, Shanghai, China;Shanghai Jiao Tong University, Shanghai, China;Shanghai Jiao Tong University, Shanghai, China;Shanghai Jiao Tong University, Shanghai, China
Venue:
Proceedings of the first workshop on Urban networking
Year:
2012

Citing 6
Cited 0

DEAPspace: transient ad hoc networking of prevasive devices

Computer Networks: The International Journal of Computer and Telecommunications Networking - pervasive computing
Jini Specification

Jini Specification
Scalable Service Discovery for MANET

PERCOM '05 Proceedings of the Third IEEE International Conference on Pervasive Computing and Communications
PDP and GSDL: A New Service Discovery Middleware to Support Spontaneous Interactions in Pervasive Systems

PERCOMW '05 Proceedings of the Third IEEE International Conference on Pervasive Computing and Communications Workshops
DSDM: A Distributed Service Discovery Model for Manets

IEEE Transactions on Parallel and Distributed Systems
Fast track article: K-directory community: Reliable service discovery in MANET

Pervasive and Mobile Computing

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper studies the Delay Bounded Service Discovery Protocol (DB-SDP) problem in Urban Vehicular Ad-hoc Networks (VANETs). Services and resources on the vehicles must be discovered before they become available to the whole network. In distributed Service Discovery Protocols (SDPs), Service Directories (SDs) are selected to store the service description for other vehicles. Selecting SDs is difficult because VANETs has the disruptive nature which incurs a significant delay, whereas the users may have QoS requirements that the query for some certain service must be answered within some delay bound. We study the problem of optimal SD selection that minimizes the number of the SDs under such delay bound requirement. We theoretically prove that such SD selection problem is NP-complete even when the future positions of the vehicles are known as a priori. To solve the DB-SDP without the prior knowledge of future traces, we theoretically and empirically analyze the number of vehicles covered by the set of SDs within a delay bound. We find the number of covered vehicles exhibits some strong regularity. Regarding this regularity, we develop a heuristic iterative algorithm for the optimal SD selection. We conduct extensive trace driven simulations based on real vehicular GPS data and the results show that with high probability, our algorithm can select SD sets 50% smaller than those selected by alternative algorithms.