Rule-based Autonomous Configuration in the AllJoyn Framework for the Internet of Things

The Internet of Things (IoT) has become a reality, thanks to the rapid growth of technologies in hardware (smartphones, wearable devices, embedded sensors, actuators), communication technologies (Wi-Fi, BLE, IEEE 802.15.4, 6LoWPAN, LTE and NFC) and software (lightweight operating systems, cloud-based services, data science). Smart and creative services such as automatic driving, energy grid, smart farming, smart home, healthcare systems and so on, are expected to become a seamless part of our daily lives. Given that various IoT-related products and services will be provided by different vendors, interoperability will be a critical issue. To solve the interoperability issue, some hardware solutions using a multi-radio gateway (such as SmartThings) is developed. However, only specific devices are supported, which is not interoperable for most of IoT products. On the other hand, there are software solutions (such as AllJoyn and IoTivity) that solve interoperability by supporting radio/platform agnostic methods and providing advertisement and discovery mechanisms. Although these software solutions address the interoperability issue at the network level, we found that heavy human involvements are still needed in order to deploy, manage and configure these devices and services. When the number of involved devices and provided services becomes larger, such a manual and usually static approach will not scale up. In addition, static configuration suffers from the lack of flexibility, especially when devices operate in a dynamic or even mobile environment. To address the scalability/flexibility issue, a rule-based configuration scheme called AutoNet is proposed. By applying this scheme, devices can (re)configure themselves to achieve an autonomous configuration in a distributed manner. These rules are constructed by some attributes including (1) the properties of devices themselves, (2) the relations between devices or (3) sensory data acquired from the environment. In order to validate the proposed scheme in real services, we implement a smart home application by applying the proposed scheme on AllJoyn. This application enables users to use smartphones to control appliances in the room where users are located. In this application, dynamic information is shared by appliances and users'' smartphones. Two metrics are evaluated to validate the performance of the proposed scheme. First, the minimal average reception period to detect an updated announcement message is 1.266 second. Second, the maximal number of supported devices to apply AutoNet on AllJoyn is 42 on average in a network. The results show that AutoNet on AllJoyn is practical and promising for various IoT applications.