Spatial big data and wireless networks: experiences, applications, and research challenges

In this article we demonstrate that spatial big data can play a key role in many emerging wireless networking applications. We also argue that spatial and spatiotemporal problems have their own very distinct role in the big data context compared to the commonly considered relational problems. We describe three major application scenarios for spatial big data, each imposing specific design and research challenges. We then present our work on developing highly scalable parallel processing frameworks for spatial data in the Hadoop framework using the MapReduce computational model. Our results show that using Hadoop enables highly scalable implementations of algorithms for common spatial data processing problems. However, development of these implementations requires significant specialized knowledge, demonstrating the need for development of more user-friendly alternatives.