Preservation and management of location privacy in the Internet of Things

The Internet of Things (IoT) connects everyday objects including a vast array of sensors, actuators, and smart devices, referred to as “things” to the Internet, in an intelligent and pervasive fashion. This connectivity gives rise to the possibility of using the tracking capabilities of things to impinge on the location privacy of users. Most of the existing management and location privacy protection solutions do not consider the low-cost and low-power requirements of things; or, they do not account for the heterogeneity, scalability, or autonomy of communications supported in the IoT. Moreover, many traditional location privacy preserving techniques anonymize location information so that adversaries cannot infer or relate location information to specific users. However, these techniques do not consider the case where a user wishes to control the granularity of the disclosed information based on the context of their use (e.g., based on the time or the current location of the user). To fill this gap, a middleware referred to as the Internet of Things Management Platform (IoT-MP) is proposed in this thesis. The IoT-MP provides users with fine-grained control over the granularity and disclosure settings of their location information in the IoT. It is based on a distributed architecture that utilises an agent, a manager, and a manager of managers paradigm. The IoT-MP adopts an extensible design where things are represented as attributes in a management database located at the manager. In this way, IoT applications can access things transparently over the Internet, irrespective of the underlying used communication technologies. The IoT-MP’s manager comprises several modules. The Privacy Module (PM), which consists of a Context Analysis Component, Privacy Manager Component, and Semantic Obfuscation Component, enables the user to alter the location of things and to control the granularity of the produced location based on a context-aware and policy enforcement mechanism. The obfuscation process is supported by a novel ontological classification of locations based on a geographical knowledge, which takes into account both the user’s informed consent and preferences. Furthermore, the proposed Semantic Obfuscation approach improves the performance of two major classical location protection methods by making it harder on an adversary to infer the actual location of a device from a received obscured location.. To confirm the effectiveness of the proposed management platform in preserving location privacy in the IoT, a diverse range of experimental and simulation studies are carried out. The experimental studies aimed to demonstrate the capability of the proposed platform in preserving the location privacy of users in an IoT setup which uses physical low-power sensor devices. The setup involved the utilisation of several Bluetooth Low Energy (BLE) sensor devices, the implementations of two mobile applications and a web application. The results collected from the experimental works validate the IoT-MP approach in providing the user with a method that can be used to control to whom, when, and in which context the location information of their sensors is revealed. They further show that the proposed Obfuscation approach has outperformed the performance of the classic Dispersion method. For instance, using “Obfuscation level 3”, it is found that the S-Obfuscation has produced better-obscured location by 60% than that of the Dispersion technique and by 50% than that of the Rand technique. The simulation studies, conducted using the Opnet and NS2 simulation tools, combined several wireless network scenarios which utilise the low-power wireless ZigBee and IEEE 802.11ah protocols as a practical example of a heterogeneous communication network in the IoT. In these scenarios, as per the IoT-MP approach, privacy policies were defined for a group of sensors which took turns in requesting the location of each other. By observing and analysing the traffic stored in the log file of the simulation, specifically, the location information exchanged between the sensors, the privacy-preserving capabilities of the proposed platform in a large-scale heterogeneous network were demonstrated and verified. Additionally, it was found that the application end-to-end delay experienced by the ZigBee network is low. Furthermore, the average consumed energy to send a packet across the network by a ZigBee and 802.11ah node was also within acceptable levels. These performance results clearly show that the approaches of the IoT-MP in preserving the location privacy of things in the IoT has no noticeable impact on the power consumptions and network performance of both ZigBee and IEEE 802.11ah end devices