Mobile networks are rapidly evolving towards new technologies that are characterized by an increasingly sophisticated radio interface, always with the aim of higher bit rates and ubiquitous coverage. Although 4G network deployments are still in their beginnings, the first upgrades towards LTE-A solutions are already planned by operators and evolutions towards 5G networks are currently under research all over the world. This rapid network evolution is essentially driven by the explosion of mobile traffic; a tendency that is not only predicted, but also observed on the actual networks. In this context, many collaborative projects are proposing new architectures and technologies as key enablers for responding to this traffic boom. Among these technologies, HetNets, composed of macro cells and all kinds of small cells, and cell coordination are regarded as the most promising ones. However, although smartphones and tablets are revolutionizing the usage of mobile networks and generating new traffic profiles, industrials, but also academics, are still evaluating and benchmarking these solutions using tools adapted to classical voice or best effort services. This raises fundamental questions about the validity of the presented performance results and their pertinence in the framework of future networks. Furthermore, technology choices within 3GPP are based on simulation results produced following the commonly agreed full buffer and FTP traffic models. While the former model does not represent any traffic reality, the latter is only adapted to best effort services. The primary aim of IDEFIX project is to radically revisit the way technologies are evaluated and benchmarked by proposing novel performance evaluation tools, based on the latest developments in queuing theory, that are able to tackle the complexity of traffic profiles in future mobile networks. These tools are to be carefully adapted to the different technologies discussed within 3GPP, and then used to benchmark these technologies and perform pertinent choices among them. Furthermore, IDEFIX will not adopt a passive behavior limited to performance evaluation of technologies. It will, on the contrary, propose service and network control mechanisms that enforce Quality of Service (QoS) and Quality of Experience (QoE) of users of different services. For this aim, this project puts together experts on performance evaluation tools and traffic engineering, whose world class research results are recognized in the telecommunication community. This expertise is complemented by another internationally recognized expertise on service and network control mechanisms and, for the first time in this field, by an expertise on network economy and decision-making in strategic investments. These academic and industrial experts will help two top actors in the world telecommunications industry, Alcatel Lucent and Orange, in their perpetual quest for producing the most efficient technologies and deploying networks with the best QoS. To summarize, the following actions are planned: - Develop a novel performance evaluation framework for future mobile networks, based on recent advances in queuing theory that considers emerging services. - Use the developed framework for dimensioning the network under target constraints on QoS/QoE for different typical traffic scenarios (service mix and traffic intensity). - Propose QoS control and enforcement mechanisms that take into account the environment fluctuations due to users’ mobility and network heterogeneity, while ensuring the lowest possible energy consumption. The most promising mechanisms will be tested using a testbed on small cells provided by ALU and proposed to 3GPP for inclusion in R12/R13. - Propose a strategic investment framework that allows the operator to plan its investments taking into account the risks related to technology maturity and traffic evolution, integrating the problem of competition/cooperation between the different actors.

The exponential surge in the global data traffic driven by the skyrocketing proliferation of different bandwidth-hungry on-line services, such as: cloud computing, on-demand HD video streams, on-line business analytics and content sharing, sensor networks, machine-to-machine traffic arising from data-centre applications, the Internet of Things, and various other broadband services, brings about the escalating pressure on the speed (capacity) and quality (bit error rate) characteristics of information systems. It is well recognized nowadays that rapidly increasing data rates in the core fibre communication systems are quickly approaching the limits of current transmission technologies, many of which were originally developed for communication over linear (e.g. radio) channels. It is widely accepted that the nonlinear transmission effects in optical fibre represent now a major limiting factor in modern fibre-optic communication systems. Nonlinear properties make optical fibre channels considerably different from wireless and other traditional linear communications channels. There is a clear need for development of radically different methods for coding, transmission, and (pre & post) processing of information that take the nonlinear properties of the optical fibre into account and for training of a new generation of engineers with expertise in: optical communications, nonlinear science methods, digital signal processing (DSP), design of implementable algorithms. From the industry perspectives, design of practical and implementable processing algorithms requires knowledge of ASICs and real world conditions and restrictions. The mutli-national & multi-interdisciplinary REAL-NET project will provide timely doctoral training for 6 PhD students through industry relevant research in the fast growing area of high practical relevance and will lead to development of novel practically implementable disruptive techniques for fibre-optic communications.

The transport industry is going through a revolutionary moment in its quest to reduce its carbon footprint, with biofuels being one of the viable options. High productivity, lipid content and its ability to capture CO2 make microalgae a competitive biomass. New technologies such as hydrothermal liquefaction (HTL) has recently attracted the attention in research as it avoids the energy-costly drying step and produces bio-oil at mild temperatures. Nonetheless, conventional HTL process suffers from bio-crudes high nitrogen content, mainly caused by protein content. Furthermore, the biocrude’s high oxygen content reduces its calorific value making biocrude unattractive in addition to its high NOx emissions during combustion. Since nitrogen and oxygen in biocrude respond differently to processing temperature, an approach that would favor simultaneous reduction of both would be revolutionary. The nitrogen removal before the main HTL process could be the solution. In this perspective, the combination of microwave pretreatment and HTL process can be promising. However, this non-selective method may compromise bio-oil yield. In MicroWCatHydroN, we investigate nitrogen-selective catalysts to specifically target nitrogen compounds and maintain the trade-off between desired bio-oil yield and properties. The catalytic activity of nickel can promote hydroisomerization and deoxygenation reactions, thereby improving bio-oil’s cold flow properties and heating value. Overall, this green chemistry-compatible route is economically and technically feasible. Industrial scale-up of algae conversion requires a complete characterization of its complex degradation mechanisms and phenomena occurring during the process in order to understand how nitrogen compounds are formed and consumed. In this sense, MicroWCatHydroN aims to develop a new chemical kinetic model for ParaChlorella kessleri liquefaction. In conclusion, research is expected to increase the production of high-quality bio-oil.

Opinion mining is a progressing domain. A lot of efforts have been recently dedicated to the development of methods able to analyze opinion data available on the social Web. At the same time, companies that are developing companion robots and virtual vocal assistants (Siri, Google Now, Cortana, etc.) show a growing interest for the integration of the social component in the interaction. MAOI is a fundamental research project in natural language processing and speech processing which contributes to Challenge 7 (Society of information and communication). It deals with opinion analysis in human-agent interaction and is thus integrated in Axis 4 (Interactions, robotics). More precisely, the MOAI project tackles multimodal opinion analysis methods in human-agent multimodal interactions in order to extract information concerning user’s preferences. Such information is dedicated to enrich user profiles for companion robots and virtual assistants. This challenging issue has been so far rarely and partially handled by the state of the art. The proposed approach relies on Conditional Random Fields (CRF) that have been chosen for their flexibility in order to take advantage of both the generalization capability of machine learning methods and the fine-grain modeling of semantic rules. As recordings of face-to face human-agent interactions are not yet massively available, such flexible methods constitute an alternative to deep learning methods. In this promising context, the MAOI project targets two major breakthroughs: i) feature learning driven by a priori knowledge and psycho-linguistic models in order to learn users' preferences; ii) the integration of various levels of analysis (lexical, syntactic, prosodic, dialogic) through latent variables inside hidden CRF, allowing for grounding the opinion detection in the context of human-agent interaction.

As Internet of Things (IoT) and IoT-Edge-Cloud continuum technologies advance, physical environments are becoming increasingly equipped with sensors, fuelling the development of smart space ecosystems. Massive quantities of data produced by IoT devices revolutionize the way such ecosystems operate via the exploitation of AI models/services. This has led to the emergence of the so-called Artificial Intelligence of Things (AIoT) systems. In general, designing techniques to promote robustness, efficiency and continual operation of AIoT systems requires realistic and trustworthy data at scale. However, such data is not always easy to obtain due to the cost of smart space construction, the inconvenience of long-term device tracking, the sensor/knowledge data gaps in diverse scenarios of a smart space, and the restrictions imposed on sensitive data sharing. Furthermore, an efficient AIoT system operation requires trustworthy AI services, as well as novel approaches for speeding up their inference across the IoT-Edge/Cloud continuum. PANDORA aims to devise and implement a comprehensive framework enabling the delivery of trustworthy datasets of smart space ecosystems, as well as the deployment and green operation of AIoT systems in such spaces. PANDORA spans two phases: (1) prior to AIoT system deployment; (2) post AIoT system deployment and operation. Phase 1 proposes and combines a series of novel techniques such as synthetic data generation, quantification of uncertainties, and data summarization for the delivery of trustworthy datasets, as well as explainable AI and domain-informed model training/testing in smart space ecosystems. Phase 2 defines novel AIaaS and CaaS techniques for the robust, explainable, green and continual operation of AIoT systems deployed in such spaces. The trustworthiness and applicability of the PANDORA framework will be tested through five pilot cases hosting AIoT applications in smart buildings, factories and critical infrastructures.