Predicting Spatio-Temporal Phenomena of Mobile Resources in Sensor Cloud Infrastructure

Abstract

Integration of sensor and cloud technologies enable distributed sensing and data collection. We consider a scenario when sensing requests are originated from sensor aware applications that are hosted inside sensor-cloud infrastructures. These requests need to be satisfied using geographically distributed sensors. However, if the sensing resources are mobile, then sensing territory is not limited to a fixed region, rather spatially diverse. In this work, we present a generic scheme for integrating spatio-temporal information of mobile sensors for Internet of Things– (IoT) based environment monitoring system. A set of algorithms are proposed in this work to model spatial and temporal features of mobile resources and exploit resource mobility. We also propose probabilistic models to measure feasibility of a resource to sense a specific spatio-temporal phenomenon. We rank the resources based on their feasibility of satisfying the sensing requests and later use the information for efficient resource allocation and scheduling.