Software Defined Radio is a programmable radio device that, when connected to a computer or as an embedded solution, can transmit and receive data information using radio waves. The programming features of the SDR and its RF bandwidth range extends the application possibility to several areas, including aviation, satellite, radar, and mobile communication. SDR has drawn great attention to network service provision. Acting as a multi-programmable air interface at the edge of wired network environments, SDR can receive, decode and forward radio information, which is used to generate the services. Examples of services including real-time flight tracker web pages, and sensor monitoring data charts. However, to provide network services, SDR must integrate into complex network environments where recent technologies, such as NFV, SDN, containerization and cloud computing, are applied. This thesis addresses the integration of SDRs with containerization. It proposes an easy-to-deploy container-based architecture to provide network services from SDR devices. Using different types of SDR devices (USRP, LimeSDR and RTL-SDR), GNURadio platform and Docker Container, two use cases of the proposed architecture are presented, demonstrating scenarios where ADS-B and LoRa communication are implemented in order to provide services to end-users. Evaluation of the proposed solution is performed comparing two models of service provision: with the proposed architecture (two levels of network isolation), and without the architecture. The overhead time added to launch the services, the time response and computational resource utilization are compared, showing that there is an overhead added by the architecture which impacts on the system performance. The overhead increases with the applied network isolation level. Conversely, the architecture converts the service functional components into modular components, its application can be extended to different RF projects and SDR types, and offers non-functional benefits such as, real-time capability, network isolation, fine setting of communication parameters, and a set of control and configuration features inherited from container virtualization platform.