A Research On The Fog Computing Based Architecture For Smart Home

With the continuous advancement of the Internet of Things technology,the smart home field has made great progress.Many traditional devices are given the functions of computing,storage,and networking.The data exchange of devices,which is realized through network interaction can provide users with diverse services.However,after the number of terminal devices is increasing and the functions are more and more complicated,the user experience is hardly to be guaranteed.Therefore,a unified smart home system is needed to manage these devices.Fog computing is an extension of cloud computing at the edge of the network.It has the characteristics of low latency,distributed,heterogeneous and interoperable,which is more suitable for smart home.Therefore,this paper proposed a lightweight architecture for smart home based on the fog computing.The architecture mainly includes two parts: the fog node and the cloud platform.The cloud platform includes three modules: image management,scenario switching analysis,and reverse proxy.The fog node is divided into the adapter layer,the data pre-processing layer and the application layer from the bottom to the top.The adapter layer considers the deployment platform of the architecture;the data pre-processing layer is responsible for managing the connected terminal devices,processing data interaction with the terminal devices,and communicate with the upper-layer applications through the message queue;the application layer runs various smart home services,which based on the data sent by the data pre-processing layer.In the traditional smart home system,the user is responsible for the management of service.When the user switches the system state,it is often necessary to control multiple services,which increases the complexity of the operation.For a better experience,this paper also proposed the concept of the scenario.The system state is corresponding to the scenario,and the scenario is corresponding to a set of services.In this way,the user only needs to operate the scenario to complete the service management.In the implementation stage,this paper adopted Docker technology to package multiple services into an image,so that the management of scenarios can be transformed into the management of container instances.With the isolation of the container technology,the architecture does not need to consider the complexity of scenario management brought by the coupling of service code.When switch to a scenario,the architecture calculates the container instances that need to be operated according to the relation between services and scenarios.This process requires a certain amount of computation.In order to ensure the lightweight property of the architecture,this paper migrates the process to the cloud platform.The local architecture sends the request to the scenario switching analysis interface of the cloud platform,and then controls the corresponding container instances according to the calculation result returned by the interface.Finally,based on the proposed lightweight architecture,this paper designed and implemented a smart home system,and evaluated the usability of architecture by function test and performance test.