Skip to main content

Resultados relacionados

Resultados totales (Incluyendo duplicados): 3
Encontrada(s) 1 página(s)

RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Artículo científico (JournalArticle)

Reconfigurable Activation Functions in Integrated Optical Neural Networks

  • Rausell-Campo, Jose Roberto|||0000-0002-6710-4766
  • Pérez-López, Daniel
[EN] The implementation of nonlinear activation functions is one of the key challenges that optical neural networks face. To the date, different approaches have been proposed, including switching to digital implementations, electro-optical or all optical. In this article, we compare the response of different electro-optic architectures where part of the input optical signal is converted into the electrical domain and used to self-phase modulate the intensity of the remaining optical signal. These architectures are made up of Mach Zehnder Interferometers (MZI) and microring resonators (MRR). We have compared the corresponding transfer functions with commonly used activation functions in state-of-the-art machine learning models and carried out an in-depth analysis of the capabilities of those architectures to generate the proposed activation functions. We demonstrate that a ring assisted MZI and a two-ring assisted MZI present the highest expressivity among the proposed structures. To the best of our knowledge, this is the first time that a quantified analysis of the capabilities of optical devices to mimic state-of-the-art activation functions is presented. The obtained activation functions are benchmarked on two machine learning examples: classification task using the Iris dataset, and image recognition using the MNIST dataset. We use complex-valued feed-forward neural networks and get test accuracies of 97% and 95% respectively., This work was supported in part by FPI-UPV Grant Program under Grant PAID-01-20-24 from the Universitat Politecnica de Valencia, through the Spanish MINECO Juan de la Cierva Program and in part by the H2020-ICT2019-2 Neoteric 871330 Project.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Artículo científico (JournalArticle)

Integrated Microwave Photonics Coherent Processor for Massive-MIMO Systems in Wireless Communications

  • Martínez-Carrasco-Romero, Pablo|||0000-0003-3771-590X
  • Rausell-Campo, Jose Roberto|||0000-0002-6710-4766
  • Pérez-Galacho, Diego
  • Li, Xu
  • Qing, Ting
  • Pérez-López, Daniel
  • Wang, Tiangxiang
[EN] Massive-MIMO systems can achieve high capacities and data rates by increasing the number of operational antennas in the base station. As more antenna elements are introduced, the complexity of the signal processing operations increases accordingly and current electronic processors struggle to reach those requirements. To overcome these hurdles, we propose a novel theory to process the RF signals in the optical domain. We experimentally demonstrate the capabilities of theory using a non-integrated Dual Parallel Modulator as a 2×2 optical matrix multiplier, recovering satisfactorily previously mixed RF signals modulated following the standards in 5G communications. Finally, we propose an integrated photonic architecture based on a core of Mach Zehnder Interferometers to physically process the information. Together with the optical core, we proposed the integration of RF modulators, microwave photonic filters and high-speed photodetectors.We present simulations on the expected performance of the circuit and analyze the impact of fabrication errors., The work of Jose Roberto Rausell-Campo was supported by the Polythecnic University of Valencia through FPI-UPV Program under Grant PAID-01-20-24. This work was supported by the UPV-Huawei Wireless Joint Innovation Center Agreement under Project YBN2020095120.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Artículo científico (JournalArticle)

Exploring Environmental Risk in Soils: Leveraging Open Data for Non-Sampling Assessment?

  • Aparisi-Navarro, Silvia
  • Candeias, Carla
  • Rocha, Fernando
  • Moncho-Santonja, María|||0000-0001-9520-6981
  • Defez Garcia, Beatriz|||0000-0002-3658-6766
  • Peris Fajarnes, Guillermo|||0000-0001-9215-9091
[EN] Soil contamination by heavy metals (HM) is a critical area of research. Traditional methods involving sample collection and lab analysis are effective but costly and time-consuming. This study explores whether geostatistical analysis with GIS and open data can provide a faster, more precise, and cost-effective alternative for HM contamination assessment without extensive sampling. Concentrations of nine HMs (Cu, Pb, Ni, Co, Mn, As, Cd, Sb, Cr) were analysed from 498 soil samples collected in two mining areas in Portugal: the Panasqueira and Aljustrel mines. Corresponding data were extracted from the Lucas TOPSOIL 1 km raster maps. Several contamination indices, Contamination Factor (Cf), Modified Contamination Degree (mCd), Geoaccumulation Index (Igeo), Nemerow Pollution Index (Pn), Potential Ecological Risk Index (PERI), and Pollution Load Index (PLI) were calculated for both datasets. A confusion matrix was used to evaluate the percentage of correct classifications, while a concordance analysis assessed the alignment of accurately classified points between the two data sources. In the soil samples, very high contamination levels for As were observed in 42% of the samples according to the Cf, with high levels for Sb found in approximately 30% of the samples. The mCd revealed that approximately 11% of soil samples exhibited very high levels of contamination, while the Pn indicated that 78.9% of the soil samples fell within the seriously polluted domain.Similar contamination trends were observed for the other indices. In contrast, the results for the LUCAS points showed significant discrepancies. No high contamination levels were found for any metal. The misclassification rates for mCd, Pn, PERI, and PLI were 84.25%, 97.55%, 95%, and 82%, respectively, when compared to the field data. This study concludes that while open data raster maps offer rapid overviews, they fall short of providing the detailed precision required for reliable contamination assessments. The significant misclassification rates observed highlight the limitations of relying solely on these tools for critical environmental decisions. Consequently, traditional sampling and laboratory analysis remain indispensable for accurate risk assessments of HM contamination, ensuring a more reliable foundation for decision-making and environmental management., This work was supported by Funding for open access charge: Universitat Polit`ecnica de Val`encia". Silvia Aparisi-Navarro was supported by the Research and Development Program (PAID-01-20) of the Universitat Politecnica de Valencia.