PhD thesis Edge-Driven Proximity Service Platform for the Internet of Things in Indoor Environments of Michael Haus
This repository addresses the RQ3 of the research problem:
How to enable user-oriented and privacy-aware IoT services by utilizing spatial proximity awareness in the managed indoor IoT areas?
We address this question by realizing a set of different proximity services. Our fine-granular seamless device association uses the similarity of visible light signaling in different spatial areas which are impossible to differentiate with Wi-Fi. On this basis, we can enable social applications, such as to retrieve knowledge from locals, e.g., Alice is a tourist, rides on the subway and wants to ask locals for the best way to the museum. Moreover, we improved the user privacy in public spaces by turning around the action cycle for service discovery to be entirely initiated from the IoT environment. The users remain passive and cannot be tracked by collecting service advertisements via distance-limited VLC when approaching a service area. Furthermore, we automate the authorization for the remote control of smart homes. A smart home incorporates a communication network that connects the key electrical appliances and services, and allows to be remotely controlled, monitored or accessed. We fully automate the key management for remote control of smart homes to improve the usability of system's security based on out-of-band VLC to exchange secret keys and a challenge-response mechanism for on-demand access requests.
For instance, as minor use case, the automated wireless authentication achieves a touchless authentication experience in a distance-bounding manner. We improve the usability by avoiding manual distribution and tedious input of passwords for login. We use LocalVLC for M2M communication to ease the setup of Wi-Fi networks. Other mechanisms to exchange credentials still require human interaction, such as WPS or QR codes. Target devices include common Wi-Fi equipped devices, e.g., smartphone, tablet, laptop, as well as IoT devices like sensor boards. The video: https://www.youtube.com/watch?v=e8kjfDNmVSA shows our working demo. Via LocalVLC we broadcast the network name (SSID) and continuously generated time-based one-time passwords (TOTPs) for the wireless login. The user's end-device retrieves the VLC transmitted login data and continually scans for nearby wireless networks. In case of spotting a matching SSID, it performs the automated authentication without any manual interaction.