Stack4Things: integrating IoT with OpenStack in a Smart City context Giovanni Merlino *‡ , Dario Bruneo * , Salvatore Distefano † , § Francesco Longo * , Antonio Puliafito * * Dipartimento DICIEAMA, Universit` a di Messina, Contrada di Dio, 98166 Messina, Italy Email: {dbruneo,flongo,gmerlino,apuliafito}@unime.it † Dipartimento DEIB, Politecnico di Milano Piazza L. Da Vinci 32, 20133 Milano, Italy Email: salvatore.distefano@polimi.it ‡ Dipartimento DIEEI, Universit` a di Catania Viale Andrea Doria 6, 98166 Catania, Italy Email: giovanni.merlino@dieei.unict.it § Machine Cognition Lab, Higher Institute of Information Technologies and Information Systems Kazan Federal University, Kazan, Russia Abstract—As the adoption of embedded systems, mobiles and other smart devices keeps rising, and the scope of their involvement broadens, for instance in the enablement of Smart City-like scenarios, a pressing need emerges to tame such complexity and reuse as much tooling as possible without resorting to vertical ad-hoc solutions, while at the same time taking into account valid options with regards to infrastructure management, and other more advanced functionalities. In this sense, a widely used and competitive framework for Infrastruc- ture as a Service such as OpenStack, with its breadth in terms of feature coverage and expanded scope, looks like fitting the bill. This work thus describes rationale, efforts, and results so far achieved, for an integration of IoT paradigms and resource ecosystems with such a kind of Cloud-oriented environment, by focusing on a Smart City scenario, and featuring data collection and visualization as example use cases of such integration. Keywords-IoT; Smart City; Cloud; IaaS; OpenStack; Ceilometer; MOM; AMQP; CoAP; REST; CEP I. I NTRODUCTION In recent years, different approaches, also based on new technologies, have been adopted in order to morph cities into smart ones. The Smart City scenario is a fertile application domain for different sciences and technologies, in particular for those related to the information and communication areas. Several ongoing projects illustrate Smart City oppor- tunities and challenges [1] in diverse application fields, e.g., networking, decision support-systems, power grids, energy- aware platforms, service-oriented architectures, highlighting the need to equip the cities of the future with a variety of urban sensors. Plenty of applications can be envisioned from traffic monitoring to energy management, from e-health to e-government, from crowd to emergency management. This all-encompassing and much ambitious scenario calls for adequate ICT technologies. In particular, solutions for managing the underlying physical sensing and actuation resources infrastructure are required. A number of solutions of this kind, mainly at a lower (communication) layer, may be found within the Internet of Things (IoT) [2] domain, mainly aiming at interconnecting network-enabled devices and, generally, any thing featuring a network interface, to the Internet. However, in order to manage devices, sensors, and things building up a dynamic Smart City infrastructure, management, organization, and coordination mechanisms are also required. In particular, Cloud computing facilities, implementing a service-oriented approach in the provisioning and management of resources, may be exploited. The Cloud approach could be a good solution to address Smart City-related issues, fitting with the requirements of relevant service users and providers: on-demand, elastic and QoS-guaranteed, to name a few, all needed properties for a Smart city service platform. Several works deal with infrastructure issues and solutions related to Smart Cities and their relationship with IoT and Cloud. Specifically, the authors of [3] propose a plat- form for managing urban services that include convenience, health, safety, and comfort. Cloud computing infrastructure recently [4], [5] found useful application in the context of Smart Cities. In [6], authors took a different trajectory to tackle transducers in a Cloud environment: sensing and actuation devices should be handled along the same lines as computing and storage abstractions in traditional Clouds, i.e. virtualized and multiplexed over (scarce) hardware re- sources. This step paves the way to an IoT/Cloud-powered Smart City framework able to support related services in the (sensing) infrastructure provisioning. Even if a lot of