Abstract
A common problem in networking research and development is the duplicate effort of writing simulation and implementation code. This duplication can be avoided through the use of fast-prototyping methodologies, which enable reusing simulation code in real prototyping and in production environments. Although this functionality is already available by using ns-3 emulation, there are still limitations regarding the support of real network interfaces and easy configuration of the network settings, such as IP and MAC addresses. In this paper we propose an improved version of the ns-3 emulation component by introducing new functionalities that address these limitations. The new functionalities include the support of new types of real network interfaces and the easier integration of emulation nodes with existing networks by means of a new auto-configuration mechanism for ns-3 nodes. Experimental results obtained in a laboratorial testbed and in a real vehicular network testbed demonstrate the new functionalities proper operation, and their backwards compatibility with previously coded ns-3 scenarios.

Abstract
Autonomous robots play a pivotal role in improving productivity and reducing operational costs. They excel at both precision and speed in repetitive jobs and can cooperate with humans in complex tasks within dynamic environments. Self-localization is critical to any robot that must navigate or manipulate the environment. To solve this problem, a modular localization system suitable for mobile manipulators was developed. By using LIDAR data the proposed system is capable of achieving less than a centimeter in translation error and less than a degree in rotation error while requiring only 5 to 25 milliseconds of processing time. The system was tested in two different robot platforms at different velocities and in several cluttered and dynamic environments. It demonstrated high accuracy while performing pose tracking and high reliability when estimating the initial pose using feature matching. No artificial landmarks are required and it is able to adjust its operation rate in order to use very few hardware resources when a mobile robot is not moving.

Abstract
The wine sector is a billion euro business and therefore subjected to multiple attempts of fraudulent practices. This requires the development of rapid and reliable methods to detect such situations. Several methodologies have been developed based on the chemical profiles of the wines, but they are limited due to the environmental conditions that cannot be controlled. The use of DNA-based detection systems are an emergent research field that have been extended to a wide variety of food products and are still the most reliable methods for varietal identification. However these methods are not suitable for geographical determination. Soil related fingerprints have a primary role considering that there is a relationship between the elemental composition of wine and the composition of the provenance soil. WineBioCode is a project aiming to define the best strategy for wine authenticity based on a multidisciplinary approach. Two DNA-based strategies have been developed based on Real-time PCR and a label free optical biosensor platform. Both platforms enabled successful identification of specific DNA-targets when applied to Vitis vinifera L., and can be applied throughout the grape-wine chain. The methods are complementary and can be used in different situations, according to the requirements. The geographical evaluation has been assessed by the strontium 875r1865r isotope ratio determination involving soil evaluation in the vineyards followed by its assay in the wine samples. The results are being integrated in order to establish the best procedure to be undertaken for wine fingerprinting, including varietal composition and geographical origin, therefore fulfilling the requirements of the geographical denominations in wine certification.

Abstract
Optimization in current decision support systems has a highly interdisciplinary nature related with the need to integrate different techniques and paradigms for solving real-world complex problems. Computing optimal solutions in many of these problems are unmanageable. Heuristic search methods are known to obtain good results in an acceptable time interval. However, parameters need to be adjusted to allow good results. In this sense, learning strategies can enhance the performance of a system, providing it with the ability to learn, for instance, the most suitable optimization technique for solving a particular class of problems, or the most suitable parameterization of a given algorithm on a given scenario. Hyper-heuristics arise in this context as efficient methodologies for selecting or generating (meta) heuristics to solve NP-hard optimization problems. This paper presents the specification of a hyper-heuristic for selecting techniques inspired in nature, for solving the problem of scheduling in manufacturing systems, based on previous experience. The proposed hyper-heuristic module uses a reinforcement learning algorithm, which enables the system with the ability to autonomously select the meta-heuristic to use in optimization process as well as the respective parameters. A computational study was carried out to evaluate the influence of the hyper-heuristics on the performance of a scheduling system. The obtained results allow to conclude about the effectiveness of the proposed approach.

Abstract

Abstract