Design and Implementation of IoT Testbed with Improved Reliability using Conditional Probability Techniques

Abstract

Internet of Things (IoT) test beds are widely used by developers predominantly. Off late, Test beds are being used by data analysts, academicians, industrial persons and hardware ardent. The real purpose of test beds, is to achieve accurate testing results, mimicking the real time environment to the extent possible which is otherwise not possible to reproduce using simulators. As predicted by many industrial giants IoT based devices will reach the scale of billions by 2015. The applications and opportunities they create will also be innumerable. This has created a huge demand for such testing grounds, because a system deployed without proper testing may be vulnerable and sometimes disastrous. Thus our research aims to explore the qualities of test beds, the services they offer and how to enhance the performance of test beds. Our test bed framework is designed and developed with open-source boards to achieved heterogeneity, reusability, interoperability and scalability. This framework would like to be addressed as a utility, with “X” as service: data, sensor client, actuator client, and platform. To achieve this, APIs which is platform and language independent has been developed and provides third-party developer support. The APIs developed have shown considerable improvement in terms of data transfer rate, database upload and retrieval, and user responsiveness. Thus, our framework is capable of offering services through our API. And have demonstrated with the help of conditional probability techniques enhancement in performance and reusability, visualized the same in terms of graphs and datasets.