Abstract
The increasing complexity of recent Wi-Fi amendments is making the use of traditional algorithms and heuristics unfeasible to address the Rate Adaptation (RA) problem. This is due to the large combination of configuration parameters along with the high variability of the wireless channel. Recently, several works have proposed the usage of Reinforcement Learning (RL) techniques to address the problem. However, the proposed solutions lack sufficient technical explanation. Also, the lack of standard frameworks enabling the reproducibility of results and the limited availability of source code, makes the fair comparison with state of the art approaches a challenge. This paper proposes a framework, named RateRL, that integrates state of the art libraries with the well-known Network Simulator 3 (ns-3) to enable the implementation and evaluation of RL-based RA algorithms. To the best of our knowledge, RateRL is the first tool available to assist researchers during the implementation, validation and evaluation phases of RL-based RA algorithms and enable the fair comparison between competing algorithms.

Abstract
Future telecommunications aim to be ubiquitous and efficient, as widely deployed connectivity will allow for a variety of edge/fog based services. Challenges are numerous, e.g., spectrum overuse, energy efficiency, latency and bandwidth, battery life and computing power of edge devices. Addressing these challenges is key to compose the backbone for the future Internet-of-Things (IoT). Among IoT applications are Indoor Positioning System and indoor Real-Time-Location-Systems systems, which are needed where GPS is unviable. The Bluetooth Low Energy (BLE) 5.1 specification introduced Direction Finding to the protocol, allowing for BLE devices with antenna arrays to derive the Angle-of-Arrival (AoA) of transmissions. Well known algorithms for AoA calculation are computationally demanding, so recent works have addressed this, since the low-cost of BLE devices may provide efficient solutions for indoor localization. In this paper, we present a system topology and algorithms for self-localization where a receiver with an antenna array utilizes the AoAs from fixed battery powered beacons to self-localize, without a centralized system or wall-power infrastructure. We conduct two main experiments using a BLE receiver of our own design. Firstly, we validate the expected behaviour in an anechoic chamber, computing the AoA with an RMSE of 10.7 degrees conduct a test in an outdoor area of 12 by 12 meters using four beacons, and present pre-processing steps prior to computing the AoAs, followed by position estimations achieving a mean absolute error of 3.6 m for 21 map positions, with a minimum as low as 1.1 m.

Abstract
The C/C++ compilation stack (Intermediate Representations (IRs), compilation passes and backends) is encumbered by a steep learning curve, which we believe can be lowered by complementing it with approaches such as source-to-source compilation. Source-to-source compilation is a technology that is widely used and quite mature in certain programming environments, such as JavaScript, but that faces a low adoption rate in others. In the particular case of C and C++ some of the identified factors include the high complexity of the languages, increased difficulty in building and maintaining C/C++ parsers, or limitations on using source code as an intermediate representation. Additionally, new technologies such as Multi-Level Intermediate Representation (MLIR) have appeared as potential competitors to source-to-source compilers at this level. In this paper, we present what we have identified as current challenges of source-to-source compilation of C and C++, as well as what we consider to be opportunities and possible directions forward. We also present several examples, implemented on top of the Clava source-to-source compiler, that use some of these ideas and techniques to raise the abstraction level of compiler research on complex compiled languages such as C or C++. The examples include automatic parallelization of for loops, high-level synthesis optimisation, hardware/software partitioning with run-time decisions, and automatic insertion of inline assembly for fast prototyping of custom instructions. © João Bispo, Nuno Paulino, and Luís Miguel Sousa.

Abstract
Advancements in semiconductor technology no longer occur at the pace the industry had been accustomed to. We have entered what is considered by many to be the post-Moore era. In order to continue scaling performance, increasingly heterogeneous architectures are being developed and the use of special purpose accelerators is on the rise. One notable example are Field-Programmable-Gate-Arrays (FPGAs), both in the data-center and embedded spaces. Advances in FPGA features and tools is allowing for critical kernels to be accelerated on specialized hardware without fabrication costs. However, re-targeting code to such heterogeneous platforms still requires significant refactoring of the compute intensive kernels, as well as knowledge of parallel compute and hardware design concepts for maximization of performance. We present Tribble, a source-to-source framework under active development, capable of transforming regular C/C++ programs for execution on heterogeneous architectures. This includes transforming the target kernel source code so that it is amenable for circuit generation while keeping the original version for software execution, inserting code for task and memory management and injecting a scheduler algorithm.

Abstract
This paper introduces two machine learning optimization algorithms to significantly enhance position estimation in Reconfigurable Intelligent Surface (RIS) aided localization for mobile user equipment in Non-Line-of-Sight conditions. Leveraging the strengths of these algorithms, we present two methods capable of achieving extremely high accuracy, reaching sub-centimeter or even sub-millimeter levels at 3.5 GHz. The simulation results highlight the potential of these approaches, showing significant improvements in indoor mobile localization. The demonstrated precision and reliability of the proposed methods offer new opportunities for practical applications in real-world scenarios, particularly in Non-Line-of-Sight indoor localization. By evaluating four optimization techniques, we determine that a combination of a Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) results in localization errors under 30 cm in 90 % of the cases, and under 5 mm for close to 85 % of cases when considering a simulated room of 10 m by 10m where two of the walls are equipped with RIS tiles. © 2023 IEEE.

Abstract
Semantic segmentation consists of classifying each pixel according to a set of classes. This process is particularly slow for high-resolution images, which are present in many applications, ranging from biomedicine to the automotive industry. In this work, we propose an algorithm targeted to segment high-resolution images based on two stages. During stage 1, a lower-resolution interpolation of the image is the input of a first neural network, whose low-resolution output is resized to the original resolution. Then, in stage 2, the probabilities resulting from stage 1 are divided into contiguous patches, with less confident ones being collected and refined by a second neural network. The main novelty of this algorithm is the aggregation of the low-resolution result from stage 1 with the high-resolution patches from stage 2. We propose the U-Net architecture segmentation, evaluated in six databases. Our method shows similar results to the baseline regarding the Dice coefficient, with fewer arithmetic operations.