The emerging communication networks tend to aggregate heterogeneous networking infrastructures as well as data flows with very distinct requisites. This implies that the complete satisfaction of Quality of Service (QoS) metrics is very difficult to achieve, using the legacy management solutions. Alternatively, the Software Defined Networking (SDN) paradigm offers a logical centralized management of the necessary network resources for data flows, namely the ones originated in sensor devices. Therefore, this work investigates a solution that meets the QoS requirements of traffic from remote Internet of Thing (IoT) devices. To achieve this goal, we have designed a SDN-based solution that manages a network topology formed by several domains. We assume each network domain is controlled by its own SDN controller. In addition, our solution assumes that the several SDN controllers need to be orchestrated among them to maximize the management efficiency of the available end-to-end network resources. This orchestration is done via an SDN transit domain ruled by the ONOS SDN-IP application. We have emulated network topologies with IoT devices to evaluate the proposed solution in terms of its functionality, robustness against network failures, and QoS support. Analyzing the obtained results, our solution can support a cross-controller SDN domain communication. It is also capable of reacting automatically to topology failures. In addition, it can prioritize the traffic within the network infrastructure, providing to the end users strong guarantees on the desired quality for the exchange of data associated to the applications they aim to use.