A hybrid scale-out cloud-based data service for worldwide sensors

We are living in a sensor-rich world. However, managing, accessing and analyzing the collective worldwide sensors' spatio-temporal observations in a coherent manner is very challenging. That is because the large number of sensors are distributed all over the world and each sensor provides large volume of continuous observations over the time. Our objective in this paper is to construct a scalable data service for gathering and accessing the worldwide sensors' collective observations. Our proposed solution has a hybrid architecture consisting of local services and a Cloud storage. In our solution, we combine a cloud-based scale out geospatial data stream architecture with the LOST-tree indexing structure. Our initial experiment shows that such hybrid structure is scalable and efficient for sensor data write, local search and global historical search.