Loading...
6 results
Search Results
Now showing 1 - 6 of 6
- Benchmarking a Wide Spectrum of Metaheuristic Techniques for the Radio Network Design ProblemPublication . Mendes, S.P.; Molina, G.; Vega-Rodriguez, M.A.; Gomez-Pulido, J.A.; Saez, Y.; Miranda, G.; Segura, C.; Alba, E.; Isasi, P.; Leon, C.; Sanchez-Perez, J.M.; Mendes, SilvioThe radio network design (RND) is an NP-hard optimization problem which consists of the maximization of the coverage of a given area while minimizing the base station deployment. Solving RND problems efficiently is relevant to many fields of application and has a direct impact in the engineering, telecommunication, scientific, and industrial areas. Numerous works can be found in the literature dealing with the RND problem, although they all suffer from the same shortfall: a noncomparable efficiency. Therefore, the aim of this paper is twofold: first, to offer a reliable RND comparison base reference in order to cover a wide algorithmic spectrum, and, second, to offer a comprehensible insight into accurate comparisons of efficiency, reliability, and swiftness of the different techniques applied to solve the RND problem. In order to achieve the first aim we propose a canonical RND problem formulation driven by two main directives: technology independence and a normalized comparison criterion. Following this, we have included an exhaustive behavior comparison between 14 different techniques. Finally, this paper indicates algorithmic trends and different patterns that can be observed through this analysis.
- The Radio Network Design Optimization ProblemPublication . Mendes, Silvio; Gómez-Pulido, Juan A.; Vega-Rodríguez, Miguel A.; Sánchez-Pérez, Juan M.; Sáez, Yago; Isasi, PedroThe fast growth and merging of communication infrastructures and services turned the planning and design of wireless networks into a very complex subject. The Radio Network Design (RND) is a NP-hard optimization problem which consists on the maximization of the coverage of a given area while minimizing the base station (BS) deployment. Solving such problems resourcefully is relevant for many fields of application and has direct impact in engineering, scientific and industrial areas. Its significance is growing due to cost dropping or profit increase allowance and can additionally be applied to several different business targets. Numerous works can be found in the literature dealing with the RND problem, although they all suffer from the same shortfall: a non-comparable efficiency. Therefore, the aim of this work is threefold: first, to offer a reliable RND benchmark reference covering a wide algorithmic spectrum, second, to offer a grand insight of accurately comparisons of efficiency, reliability and swiftness of the different employed algorithmic models and third, to disclose reproducibility details of the implemented models, including simulations of a hardware co-processing accelerator.
- Driving Behavior Classification Using a ConvLSTMPublication . Pingo, Alberto; Castro, João; Loureiro, Paulo; Mendes, Silvio; Bernardino, Anabela; Miragaia, Rolando; Husyeva, IrynaThis work explores the classification of driving behaviors using a hybrid deep learning model that combines Convolutional Neural Networks (CNNs) with Long Short-Term Memory (LSTM) networks (ConvLSTM). Sensor data are collected from a smartphone application and undergo a preprocessing pipeline, including data normalization, labeling, and feature extraction, to enhance the model’s performance. By capturing temporal and spatial dependencies within driving patterns, the proposed ConvLSTM model effectively differentiates between normal and aggressive driving behaviors. The model is trained and evaluated against traditional stacked LSTM and Bidirectional LSTM (BiLSTM) architectures, demonstrating superior accuracy and robustness. Experimental results confirm that the preprocessing techniques improve classification performance, ensuring high reliability in driving behavior recognition. The novelty of this work lies in a simple data preprocessing methodology combined with the specific application scenario. By enhancing data quality before feeding it into the AI model, we improve classification accuracy and robustness. The proposed framework not only optimizes model performance but also demonstrates practical feasibility, making it a strong candidate for real-world deployment.
- On the Use of Perfect Sequences and Genetic Algorithms for Estimating the Indoor Location of Wireless SensorsPublication . Ferreira, Marco; Bagarić, J.; Lanza-Gutierrez, Jose M.; Mendes, Silvio; Pereira, João; Gomez-Pulido, Juan A.Determining the indoor location is usually performed by using several sensors. Some of these sensors are fixed to a known location and either transmit or receive information that allows other sensors to estimate their own locations. The estimation of the location can use information such as the time-of-arrival of the transmitted signals, or the received signal strength, among others. Major problems of indoor location include the interferences caused by the many obstacles in such cases, causing among others the signal multipath problem and the variation of the signal strength due to the many transmission media in the path from the emitter to the receiver. In this paper, the creation and usage of perfect sequences that eliminate the signal multipath problem are presented. It also shows the influence of the positioning of the fixed sensors to the precision of the location estimation. Finally, genetic algorithms were used for searching the optimal location of these fixed sensors, therefore minimizing the location estimation error.
- Planning the Deployment of Indoor Wireless Sensor Networks Through Multiobjective Evolutionary TechniquesPublication . Lanza-Gutierrez, Jose M.; Gomez-Pulido, Juan A.; Mendes, Silvio; M. Ferreira; Pereira, J. S.This work deals with how to efficiently deploy an indoor wireless sensor network, assuming a novel approach in which we try to leverage existing infrastructure. Thus, given a set of low-cost sensors, which can be plugged into the grid or powered by batteries, a collector node, and a building plan, including walls and plugs, the purpose is to deploy the sensors optimising three conflicting objectives: average coverage, average energy cost, and average reliability. Two MultiObjective (MO) genetic algorithms are assumed to solve this issue, NSGA-II and SPEA2. These metaheuristics are applied to solve the problem using a freely available data set. The results obtained are analysed considering two MO quality metrics: hypervolume and set coverage. After applying a statistical methodology widely accepted, we conclude that SPEA2 provides the best performance on average considering such data set.
- Accelerating floating-point fitness functions in evolutionary algorithms: a FPGA-CPU-GPU performance comparisonPublication . Gomez-Pulido, Juan A.; Vega-Rodriguez, Miguel A.; Sanchez-Perez, Juan M.; Priem-Mendes, Silvio; Carreira, VitorMany large combinatorial optimization problems tackled with evolutionary algorithms often require very high computational times, usually due to the fitness evaluation. This fact forces programmers to use clusters of computers, a computational solution very useful for running applications of intensive calculus but having a high acquisition price and operation cost, mainly due to the Central Processing Unit (CPU) power consumption and refrigeration devices. A low-cost and high-performance alternative comes from reconfigurable computing, a hardware technology based on Field Programmable Gate Array devices (FPGAs). The main objective of the work presented in this paper is to compare implementations on FPGAs and CPUs of different fitness functions in evolutionary algorithms in order to study the performance of the floating-point arithmetic in FPGAs and CPUs that is often present in the optimization problems tackled by these algorithms. We have taken advantage of the parallelism at chip-level of FPGAs pursuing the acceleration of the fitness functions (and consequently, of the evolutionary algorithms) and showing the parallel scalability to reach low cost, low power and high performance computational solutions based on FPGA. Finally, the recent popularity of GPUs as computational units has moved us to introduce these devices in our performance comparisons. We analyze performance in terms of computation times and economic cost.