Abstract

The efficient data access of streaming vehicle data is the foundation of analyzing, using and mining vehicle data in smart cities, which is an approach to understand traffic environments. However, the number of vehicles in urban cities has grown rapidly, reaching hundreds of thousands in number. Accessing the mass streaming data of vehicles is hard and takes a long time due to limited computation capability and backward modes. We propose an efficient streaming spatio-temporal data access based on Apache Storm (ESDAS) to achieve real-time streaming data access and data cleaning. As a popular streaming data processing tool, Apache Storm can be applied to streaming mass data access and real time data cleaning. By designing the Spout/bolt workflow of topology in ESDAS and by developing the speeding bolt and other bolts, Apache Storm can achieve the prospective aim. In our experiments, Taiyuan BeiDou bus location data is selected as the mass spatio-temporal data source. In the experiments, the data access results with different bolts are shown in map form, and the filtered buses’ aggregation forms are different. In terms of performance evaluation, the consumption time in ESDAS for ten thousand records per second for a speeding bolt is approximately 300 milliseconds, and that for MongoDB is approximately 1300 milliseconds. The efficiency of ESDAS is approximately three times higher than that of MongoDB.

Highlights

  • IntroductionVehicle observation streaming mass data access is a research hotspot in smart cities [1,2,3]

  • Vehicle observation streaming mass data access is a research hotspot in smart cities [1,2,3].When the size of vehicles is n, the frequency of streaming data access is n per second

  • Considering the characteristics of spatio-temporal streaming data, we proposed efficient streaming spatio-temporal data access based on Apache Storm (ESDAS) to achieve real-time streaming mass spatio-temporal vehicle data access and customized, flexible, and multi-level data cleaning

Read more

Summary

Introduction

Vehicle observation streaming mass data access is a research hotspot in smart cities [1,2,3]. When the size of vehicles is n, the frequency of streaming data access is n per second. With the rapid increase in the number of vehicles in urban cities, the frequency of data access is raised exponentially. Sensor networks produce continuous real-time and time-ordered data, called streaming data [4]. As analyzed by Choong, the characteristics of streaming data are unbounded data size, structured or semi-structured, time ordered and an unfixed data unit [5]. Accessing the streaming data is difficult, so the efficient real-time processing of data is hard to achieve [6]. As the number of sensors increases, the efficient real-time processing of data becomes more difficult [7]

Methods
Results
Conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call