Few studies have investigated and proposed a middleware solution for the Internet of Mobile Things (IoMT), where the smart things (Smart Objects) can be moved, or else can move autonomously, and yet remain accessible and controllable remotely from any other computer over the Internet. Examples of mobile Smart Objects include vehicles of any kind, wearable devices, sensor tags, mobile robots, etc, anyhing with embedded sensors and/or actuators. In this context of general mobility of objects, mobile personal devices (smart phones, tablets, etc.) are well suited as the universal providers of Internet connectivity and location information for simpler smart objects that lack location sensors and have only short-range wireless interfaces. This paper describes Mobile Hub (M-Hub), a generic mobile middleware for IoMT, its design and prototype implementation for Android and Bluetooth. The Mobile Hub extends our previous mobile-cloud communication middleware SDDL, so that it is able to provide scalable and reliable mobile communication and data processing capabilities to mobile smart objects. Preliminary experiments have shown that our implementation of M-Hub delivers good mobility responsiveness and that the concept is suitable for IoT applications that require opportunistic discovery and connection to a variety of mobile Smart Objects.

The current ubiquity of smart phones with mobile Internet and several short-range wireless interfaces (NFC, Bluetooth, Bluetooth Smart) and the fact that these devices are carried almost anytime and anywhere by users, enables potentially new pervasive sensing applications where smartphones can act as universal hubs for interaction with sensors (or sensor networks) that have only short-range wireless connectivity. Thus, in next years we can expect an increasing number of long-term and large-scale deployments for various crowd-sourced monitoring applications, such as environment monitoring, domestic utility meter reading, urban monitoring, etc. In this paper, we present the implementation and initial performance results with our mobile-cloud middleware that enables such opportunistic mobile sensing. One of the singular features of our middleware is the capability to discover, dynamically download and install sensor-specific transcoding modules on the mobile phone according to the encountered sensor type and make.

No abstract