MapReduce Processing on IoT Clouds

We present a data processing framework for IoTs with clouds. It was inspired by the MapReduce processing but aims at data processing at the edges through low-band networks. It enables data at the edges to be locally processed as much as possible by using the MapReduce processing and to be cooperated with cloud computing if the edges lacks computational capabilities. This paper describes the design and implementation of the framework, basic performance, and an practical application.