New Secure IoT Architectures, Communication Protocols and User Interaction Technologies for Home Automation, Industrial and Smart Environments

Abstract

The Internet of Things (IoT) presents a communication network where heterogeneous physical devices such as vehicles, homes, urban infrastructures or industrial machinery are interconnected and share data. For these communications to be successful, it is necessary to integrate and embed electronic devices that allow for obtaining environmental information (sensors), for performing physical actuations (actuators) as well as for sending and receiving data (network interfaces). This integration of embedded systems poses several challenges. It is needed for these devices to present very low power consumption. In many cases IoT nodes are powered by batteries or constrained power supplies. Moreover, the great amount of devices needed in an IoT network makes power e ciency one of the major concerns of these deployments, due to the cost and environmental impact of the energy consumption. This need for low energy consumption is demanded by resource constrained devices, con icting with the second major concern of IoT: security and data privacy. There are critical urban and industrial systems, such as tra c management, water supply, maritime control, railway control or high risk industrial manufacturing systems such as oil re neries that will obtain great bene ts from IoT deployments, for which non-authorized access can posse severe risks for public safety. On the other hand, both these public systems and the ones deployed on private environments (homes, working places, malls) present a risk for the privacy and security of their users. These IoT deployments need advanced security mechanisms, both to prevent access to the devices and to protect the data exchanged by them. As a consequence, it is needed to improve two main aspects: energy e ciency of IoT devices and the use of lightweight security mechanisms that can be implemented by these resource constrained devices but at the same time guarantee a fair degree of security. The huge amount of data transmitted by this type of networks also presents another challenge. There are big data systems capable of processing large amounts of data, but with IoT the granularity and dispersion of the generated information presents a new scenario very di erent from the one existing nowadays. Forecasts anticipate that there will be a growth from the 15 billion installed devices in 2015 to more than 75 billion devices in 2025. Moreover, there will be much more services exploiting the data produced by these networks, meaning the resulting tra c will be even higher. The information must not only be processed in real time, but data mining processes will have to be performed to historical data. The main goal of this Ph.D. thesis is to analyze each one of the previously described challenges and to provide solutions that allow for an adequate adoption of IoT in Industrial, domestic and, in general, any scenario that can obtain any bene t from the interconnection and exibility that IoT brings.