The Internet of Things (IoT) is the combination of various technologies and existing research fields and, with it, new disciplines, new applications' areas, and new challenges arose. One of the main fields of interest is given by Wireless Sensors Networks (WSNs), from which various applications (e.g., home automation, optimization of production processes) and various research fields (e.g., routing protocols, wireless technologies, energy optimization) derive. Regarding research and development, with the increase and adoption of the IoT paradigm, new communication protocols have been developed. In order to meet the requirements of typical IoT applications, such as low power consumption and long distance communications, the following new class of networks has been defined: Low Power Wide Area Networks (LPWANs). LPWANs are formed by very constrained nodes, especially in terms of computation capabilities and data rate transmission. However, the LPWAN nodes have very limited power consumption and very long transmission range (on the order of tens of kilometers) and the end nodes do not support Internet Protocol (IP). In the past years, moreover, many organizations, such as the Internet Protocol for Smart Objects (IPSO) Alliance and the Internet Engineering Task Force (IETF), focused their work on the design of an IoT IP-based protocol stack which can match the stack of the Internet, with particular attention to recently emerging LPWANs. In fact, with the goal of achieving interaction between heterogeneous devices, the use of IP has been widely accepted as the driver for the effective evolution of the IoT and its integration with the Internet. In this thesis, we thus propose a new hybrid architecture that enables the interaction between low-power, non-IP, and long-range devices (i.e., LoRaWAN devices) and shorter range, IP-based, IEEE 802.11-based devices. In this way, the potentialities of micro IoT networks is extended with the long range feature, creating a highly scalable IoT architecture which allows to better address the complexity of the requirements of IoT scenarios.