The Vehicular Sensor Network (VSN) technology empowers Intelligent Transportation Systems (ITS) to support a wide range of road safety and traffic management applications. By taking advantage of information collection and communication capabilities offered by VSNs, information, such as speed, travel time, dash-camera video, etc., can be gathered from sensors embedded in vehicles, and then delivered to ITS applications. Based on the data collected from VSNs, ITS applications are able to offer different kinds of services. Furthermore, high quality data is of great importance for ITS applications to offer high quality services. Quality of Information (QoI) which is a set of attributes including cost, accuracy and timeliness, is widely used to evaluate the quality of the collected data. To provide high quality data to ITS applications, we propose a Quality-oriented Data Collection (QDC) scheme. QDC aims to effectively support the accuracy and real-time requirements stipulated by ITS applications while reducing bandwidth consumption. In addition, to make use of technologies such as big data analytics, cloud computing, etc., many ITS applications adopt the centralized approach in which data is first collected from vehicles and then updated to a central server located at a remote location. The central server is able to get a global view of the status of the road by analyzing the collected data and provide valuable services to ITS application users. However, to satisfy the data collection requirements of different ITS applications, a huge number of packets are generated, which may exhaust the available wireless communication bandwidth and overload the central server. To alleviate the workload of the central server as well as to reduce the bandwidth consumption, we adopt the basic idea behind the MapReduce framework to support large-scale analytical processing in VSNs. Specifically, MapReduce supports large-scale analytical processing workloads by (i) utilizing a cluster of “data nodes” as a distributed storage; (ii) scheduling data processing tasks as close to where the data is located as possible. We propose an analytical processing framework for VSNs called Vehdoop. Vehdoop utilizes the computing capability of vehicles to efficiently process sensor data in parallel across a large number of vehicles in a decentralized manner. Furthermore, Vehdoop efficiently decreases the bandwidth consumption by adopting the data aggregation pattern of MapReduce. We conducted extensive experiments using vehicle trajectories generated by SUMO on real-world road maps and a network simulator, NS-3 for simulating Vehicle-to-Vehicle (V2V) and Vehicle-to-Infrastructure (V2I) wireless communications in VSNs. The experimental results demonstrate the superiority and effectiveness of QDC and Vehdoop.
Supporting Quality-oriented Data Collection and Data Analytical Processing in Vehicular Sensor Networks
NIE, W. (Author). 31 Aug 2018
Student thesis: Doctoral Thesis