摘要

Nowadays, more and more researchers focus on the Internet of Things (IoT), which blurs the line between virtual and real worlds. However, few of them concern how to realize this kind of applications. In order to do some further study on the software implementation, in this paper, we propose a model of Internet of things used in agriculture, named AGIOT. A new four-layer architecture of the IoT application is established and some key issues in system implementation are discussed. Further more, a new middle layer framework is presented based on service oriented architecture (SOA), which makes users access to real-time data easily through a high-level abstract interface based on web services. In order to increase the yield and improve the quality of plant, an intelligent algorithm based on reinforcement learning is introduced. The model we proposed is feasible.