On the Design of a Flexible Software Platform for In-Building OTT Service Provisioning Tim De Pauw *† , Bruno Volckaert * , Filip De Turck * and Veerle Ongenae † * Department of Information Technology (INTEC) Ghent University – IBBT, Gaston Crommenlaan 8 bus 201, 9050 Ghent, Belgium † Faculty of Applied Engineering Sciences (INWE) University College Ghent, Schoonmeersstraat 52, 9000 Ghent, Belgium Email: tim.depauw@intec.ugent.be Abstract—We propose a software platform which pairs context awareness with over-the-top (OTT) service deployment. By aug- menting OTT services with local context information, we allow them to react upon various types of changes in the environment in which they are being deployed. This lets service providers offer more personalized and fine-grained applications, while making use of a third-party infrastructure, via the OTT paradigm. Through UML diagrams, we describe the architecture of the proposed service platform. By means of a detailed illustrative sce- nario, the components involved are further clarified. In addition, in order to prove the feasibility of the architecture, a prototype implementation was developed and deployed on a large wireless sensor network test bed. Using a set of benchmarks, we identified the strengths and weaknesses of both test bed and prototype. I. I NTRODUCTION Recently, the introduction of so-called over-the-top (OTT) service providers has given rise to new business models for network-enabled services. OTT providers collaborate with network operators in order to tune the services at hand to the network infrastructure. The most common application of the OTT model is without a doubt found in the digital video domain, where content providers rely on network operators to provide often interactive video content to set-top boxes, television sets and personal computers. In addition to this classic example of the over-the-top model, the flexibility introduced by its loosely collaborative nature is likely to drive a variety of services backing the future Internet. One factor in such up-and-coming network-based services is context awareness, i.e. the ability for an application to intelligently adapt to its changing environment. In order for a given application to adapt to the environ- ment in which it is operating, context information must be readily accessible. One way of collecting such information is by processing data from hardware sensors. In a wireless sensor network (WSN), sensor equipment is generally placed alongside more traditional machines, to collect relevant data. Once raw data has been collected, knowledge must be extracted from it, so appropriate action can be taken. Often- times, this operation is performed via ontological reasoning. Ontologies allow for the specification of a semantic model of the problem domain. By consequently applying a reasoner to an ontology, additional information is inferred from the model. The WiLab [1] experimental test bed is a typical WSN. It is comprised of several hundred wireless mesh network nodes, the majority of which are equipped with sensors measuring temperature, humidity, and light intensity. The nodes are embedded systems, modest yet capable of a fair amount of local processing. Thus, the individual nodes’ resources can be harvested to drive a distributed context-aware application running on the WSN. In this paper, we present a software architecture for the operation of context-aware services in WSNs, using the over- the-top paradigm. We also describe a deployment scenario and use it to detail the purpose of the components we introduce. In addition, we list the results of a set of benchmarks, which were used to obtain performance metrics of key components of our platform. These metrics were subsequently used as parameters in resource provisioning algorithms we have developed. The remainder of this paper is structured as follows. Sec- tion II provides an overview of related work. In Section III, we describe the building blocks of our context-aware OTT service architecture, after which Section IV lists our technol- ogy choices for its implementation. Section V introduces our evaluation scenario and illustrates how the platform’s architec- ture’s components interact. The supporting benchmarks from the used experimental facilities follow in Section VI. We end with a look at opportunities for future research. II. RELATED WORK Context awareness and ubiquitous computing are currently important research topics. Ailisto et al. introduce a five-layered model for context-aware applications [2]. Our platform builds upon this model, extending it to support the OTT paradigm. Ahmad and Begen predict that next-generation networks will be mostly driven by demand for over-the-top video services. An important aspect of these so-called medianets will be targeted advertising. As digital media environments become more and more aware of their users and surroundings, content providers will be able to identify micro-scale advertising opportunities. Thus, sophisticated monetization schemes are likely to emerge in the not so distant future. [3] Ontological reasoning is often used in context-aware appli- cations, as ontologies support specification and processing of context information independent from other application logic.