Smart and green buildings: a bottom up approach

Can we develop smart homes and offices which have a sustainable impact on the environment? This crucial question has been raised by the recent dramatic rise in the diffusion of wireless IoT devices, coupled with the difficulty in disposing of exhausted batteries. In this thesis we investigated the technologies and the solutions that in the next years will populate our houses and offices, selecting and working with the most sustainable ones, paying attention to the users and to their security. Our investigation, focused on the network side, started from the physical layer up to the application one, following the TCP/IP stack, and went through a large number of technologies, from machine learning to blockchain. We worked on a battery free controller, named JoyTag; a radio fingerprint machine learning model; a reinforcement learning based MAC protocol, APT-MAC; an enhanced MAC protocol with Q-Learning and real time scheduling, ReLEDF; an SDN-based smart building management software, SECY; a smart building software emulator, SMARTEEX; and a blockchain based framework for smart building security, called HyBloSe.