Abstract
Over the last decades web and mobile technologies are increasingly being used in sports, especially in soccer, but none of them seems to allow to prevent injuries. However, training systems for young athletes do not have, for the most part, learning abilities in order to adapt, evolve and find new training recommendations. It is in this context which the Smart Coach project is presented in this work, and whose main goal is to introduce our mobile training recommendation system allow to young athletes evolve. The training mobile recommendation system is also designed to identify potential injuries risk for each young athlete.

Abstract
Low Power Wide Area Networks (LPWAN) enable a growing number of Internet-of-Things (IoT) applications with large geographical coverage, low bit -rate, and long lifetime requirements. LoRa (Long Range) is a well-known LPWAN technology that uses a proprietary Chirp Spread Spectrum (CSS) physical layer, while the upper layers are defined by an open standard - LoRaWAN. In this paper, we propose a simple yet effective method to improve the Quality-of-Service (QoS) of LoRa networks by fine-tuning specific radio parameters. Through a Mixed Integer Linear Programming (MILP) problem formulation, we find optimal settings for the Spreading Factor (SF) and Carrier Frequency (CF) radio parameters, considering the network traffic specifications as a whole, to improve the Data Extraction Rate (DER) and to reduce the packet collision rate and the energy consumption in LoRa networks. The effectiveness of the optimization procedure is demonstrated by simulations, using LoRaSim for different network scales. In relation to the traditional LoRa radio parameter assignment policies, our solution leads to an average increase of 6% in DER, and a number of collisions 13 times smaller. In comparison to networks with dynamic radio parameter assignment policies, there is an increase of 5% and 2% of DER, and a number of collisions 11 and 2.5 times smaller than equal-distribution, and random distribution, respectively.

Abstract
An electronic business transaction among untrusted bodies without consulting a mutually trusted party has remained widely accepted problem. Blockchain resolves this problem by introducing peer-to-peer network with a consensus algorithm and trusted ledger. Blockchain originally introduced for cryptocurrency that came with proof-of-work consensus algorithm. Due to some performance issues, scientists brought concept of permissioned Blockchain. Hyperledger Fabric is a permissioned Blockchain targeting business-oriented problems for industry. It is designed for efficient transaction execution over Blockchain with pluggable consensus model; however, there is limitation of rapid application development. Hyperledger introduced a new layer called Hyperledger Composer on top of the Fabric layer, which provides an abstract layer to model the business application readily and quickly. Composer provides a smart contract to extend the functionality and flexibility of Fabric layer and provides a way of communication with other systems to meet business requirements. Hyperledger Composer uses role-based access control (RBAC) model to secure access to its valuable assets. However, RBAC is not enough because many business deals require continuous assets monitoring. Our proposed model, BlockU, covers all possible access control models required by a business. BlockU can monitor assets continuously during transactions and updates attributes accordingly. Moreover, we incorporate hooks in Hyperledger Composer to implement extended permission model that provides extensive permission management capability on an asset. Subsequently, our proposed enhanced access control model is implemented with a minimal change to existing Composer code base and is backward compatible with the current security mechanism.

Abstract
The integration of new technologies at the residential level such as energy storage systems, electric vehicles, solar photovoltaic generation and mini wind turbines triggered the appearance of a new agent in the power systems called prosumers. This agent has the potential to provide new forms of flexibility and cost-effective solutions. However, associated with these new solutions there are also a number of problems that affect these solutions, particularly network constraints. This work presents an analysis not only on the benefits of utilizing the prosumer's flexibility but also to the problems associated with the operation and optimization of the network. A new model is presented that considers energy transactions between prosumers in the neighborhood and between them and the network using on a stochastic framework, in order to account for a set of uncertainties in the form of scenarios associated with the availability of various resources and technologies. The results show the economic benefit of energy transactions between prosumers resulting in more flexibility for the system while highlighting the effect of network restrictions and potential problems associated with them.

Abstract
Recently Shannon’s Entropy has been incorporated in nature inspired metaheuristics with good results. Depending on the problem, the Grey Wolf Optimization (GWO) algorithm may suffer from premature convergence. Here, an Entropy Grey Wolf Optimization (E-GWO) technique is proposed with the overall aim to improve the original GWO performance. The entropy is used to track the GWO swarm diversity, comparing the distance values between the Alpha in relation to the Beta and Delta wolves. The aim of the E-GWO variant is to improve convergence and prevent stagnation in local optima, since ideally restarting the swarm agents will prevent this from happening. Simulation results are presented showing that E-GWO restarting mechanism can achieve better results than the original GWO algorithm for some benchmark functions.

Abstract
This work seeks to design and implement a humanoid robotic kick for situations where the robot is moving for the RoboCup simulation 3D robotic soccer league. It employs Reinforcement Learning (RL) techniques, namely the Proximal Policy Optimization (PPO) algorithm to create fast and reliable skills. The kick was divided into 6 cases according to initial conditions and separately trained for each of the cases. A series of kicks, both static and in motion, using two different gaits were developed. The kicks obtained show very high reliability and, when compared to state of the art kicks, displayed a very high time performance improvement. This opens the door to more dynamic games with faster kicks in the RoboCup simulation 3D league.