Design And Implementation Of Measurement-Driven IoT Emulation Experiment Mechanisms

The Internet of Things is a technology derived from the development of the Internet and sensor network technology.It is an indispensable component of modern information technology.Many of the visions of using the Internet of Things to improve the quality of human life are being more and more realized.Before an IoT product or system is deployed to a target environment,they need to be tested and evaluated experimentally.However,there is not a good Internet of Things experiment platform to meet the experimenter's requirements for the fidelity of the experiment environment,the scalability of the experiment scale,the convenience of the experiment configuration.Therefore,the measurement-driven method is used to improve the experiment mechanism of the Internet of Things emulation,and to explore solutions to the above problems.In this paper,a method of measurement-driven and emulation is used to design and implement an IoT experiment mechanism with high fidelity,scalability and convenience.Firstly,the measurement-driven method is introduced into the Internet of Things emulation experiment,and the actual sensor data is used as the driving method.The following three experiment mechanisms are designed: based on the recommended IoT emulation experiment mechanism,based on the recommended and customized IoT emulation experiment.Mechanism based on user-defined IoT emulation experiment mechanism.Secondly,under the user-defined emulation experiment,the pseudo-hexagonal cellular grid is used for emulation nodedeployment,and the coordinate-based data mapping algorithm is proposed to distribute the sensing data of the real node to the emulation node.In addition,in order to satisfy the different sampling frequencies of the emulation nodes,an interpolation algorithm is used to supplement the time continuity of the data.Finally,the emulation experiment sensor data configuration tool is developed to realize the configuration of the IoT emulation experiment and the sensor data management,optimize the emulation experiment process,and improve the convenience of the experiment configuration.In order to evaluate the experiment mechanism of this paper,experiments were conducted separately for the three research objectives.The error evaluation of the sensor data is carried out by cross-validation,and the cost of sensor data is evaluated for the query time and node energy consumption of the emulation node.The time series model ARIMA is used to evaluate the authenticity of the model based regression.The evaluation results show that the experiment mechanism of this paper improves the authenticity of the emulation experiment while ensuring high scalability,and brings more convenience and flexibility to the experiment configuration.