XXXVIII Simpósio Brasileiro de Telecomunicações e Processamento de Sinais

  Página de 26  

A Block-Sparsity Induced NLMS Algorithm with Bias Compensation
Rodrigo M.S. Pimenta, Leonardo Resende, Lucas da Silva, Newton Siqueira, Jurair Rosa, Mariane R Petraglia, Diego B. Haddad

DOI: 10.14209/SBRT.2020.1570658250
Keywords: Block-Sparsity Bias Compensation Selective Partial Update
Abstract
Adaptive filtering algorithms are flexible mechanisms that adapt themselves to the environment statistics in which they are immersed. It is known that in practice several transfer functions are sparse, in the sense that their energy is concentrated in a few (sometimes clustered) coefficients. In this paper, a new normalized adaptive algorithm tailored to identifying block-sparse systems using a mixed ℓ2,0-norm of the adaptive coefficients is devised. Since the presence of noise in the input signal may induce an additional asymptotic bias in the estimation procedure, a compensation scheme is also advanced to address such an issue. At last, the computational burden is controlled by the adoption of a selective partial-update strategy. Simulated results indicate that the proposed algorithms present good performance compared to state-of-the-art alternatives, and allows the designer the choice of a convenient point regarding the trade-off between computational cost and convergence rate.

Download
Um Algoritmo Rápido para a Multiplicação de Quatérnios
Luciano Barboza Silva, Hélder Barbosa, Gilson Jerônimo da Silva Jr., Ricardo M Campello de Souza

DOI: 10.14209/SBRT.2020.1570658256
Keywords: Quaternions Fast algorithms Self-similar matrices Hypercomplex numbers
Abstract
This work proposes a fast algorithm for computing a product of two quaternions. The direct computation of this product requires 16 multiplications and 12 additions. The algorithm proposed here uses 8 multiplications and 28 additions, thus reducing the multiplicative complexity for computing the quaternion product.

Download
Um Algoritmo Rápido para a Multiplicação de Octônios
Luciano Barboza Silva, Hélder Barbosa, Gilson Jerônimo da Silva Jr., Ricardo M Campello de Souza

DOI: 10.14209/SBRT.2020.1570658261
Keywords: Octonions Fast algorithms Self-similar matrices Hypercomplex numbers
Abstract
This work proposes a fast algorithm for computing a product of two octonions. The direct computation of an octonion product requires 64 multiplications and 56 additions, while the proposed algorithm uses 30 multiplications and 94 additions, reducing the multiplicative complexity necessary to compute this product.

Download
Equalizador LRA-MMSE para sistemas OQ-OFDM de baixa resolução
Gabriel Silva, João T Dias

DOI: 10.14209/SBRT.2020.1570648841
Keywords: LRA-MMSE OQ-OFDM Conversores AD
Abstract
Os sistemas OFDM (Orthogonal Frequency Division Multiplexing) têm requisito de alta resolução dos conversores Analógico-Digitais, e o consumo de energia do sistema é proporcional à resolução desses conversores. Neste trabalho, derivamos um equalizador baseado no mínimo erro quadrático médio que leva em consideração a resolução do quantizador (LRA-MMSE) capaz de mitigar o efeito do ruído de quantização no domínio da frequência em sistemas OFDM sobreamostrados e quantizados (OQ-OFDM) com baixa resolução. Simulações revelam que o modelo proposto resulta em ganho de taxa de erro de bit, reduzido consumo de energia e melhor soma das taxas que os receptores OFDM tradicionais.

Download
Modelo Complexo eta-mu Bivariável com Desbalanceamento de Clusters
Alessandro Paulo de Oliveira, Thiago Bairros, Rausley Adriano Amaral de Souza, Michel Daoud Yacoub

DOI: 10.14209/SBRT.2020.1570658277
Keywords: eta-mu distribution fading channels bivariate phase envelope joint distribution.
Abstract
In this paper, a fading model for the complex bivariate eta-mu channel, with an imbalance between the number of clusters of the in phase and quadrature components, is proposed. An exact closed-form expression for the joint probability density function including the two envelopes and their corresponding phases is obtained. It is assumed that the correlation occurs between the respective in phase components with each other and the respective in quadrature components with each other of the eta-mu signals. The proposed model generalizes two models already presented elsewhere in the literature.

Download
Mapa de Saliência para Nuvem de Pontos usando Projeções
Victor F Figueiredo, Ricardo de Queiroz

DOI: 10.14209/SBRT.2020.1570658395
Keywords: Mapa de saliência Nuvem de pontos
Abstract
Algorithms for creating saliency maps are well established for images, even though there is no literature on such methods for point clouds. We use orthographic projections in 2D planes which are subject to well established saliency detection algorithms to create a 3D saliency map. The results of each saliency map are projected to the 3D voxels and the results of the many projections are used to generate a 3D saliency map. As a future work it is proposed to use the saliency map to assist in the compression of point cloud.

Download
Digital Nonlinearity Compensation Techniques for Unrepeatered Optical Systems
José Hélio da Cruz Júnior, Tiago Sutili, Lucas Silva Schanner, Sandro M. Rossi, Rafael C. Figueiredo

DOI: 10.14209/SBRT.2020.1570658398
Keywords: Digital back-propagation Maximum likelihood sequence estimation Nonlinear compensation Unrepeatered systems
Abstract
In this paper, we experimentally investigate the performance of unrepeatered optical transmission comparing different nonlinear compensation (NLC) implementations. Specifically, digital back-propagation (DBP) algorithm and maximum likelihood sequence estimation (MLSE) are applied for intra-channel NLC with and without 4x4 multiple-input multiple-output (MIMO) equalization. Both NLC algorithms are evaluated in an unrepeatered WDM transmission of 17x200-Gb/s channels (32 GBd DP-16QAM) over 350 km of large effective area and low loss single-mode fibers. The results indicate a Q^2 factor and optimum launch power improvements of up to 0.4 dB and 2 dB, respectively, compared to the linear compensation (LC), combining MLSE and DBP with MIMO equalization.

Download
Análise de Robustez de um Método de Redução de Ruído para Implantes Cocleares
Rafael Attili Chiea, Marcio H Costa

DOI: 10.14209/SBRT.2020.1570644875
Keywords: Redução de ruído Robustez Implante Coclear
Abstract
Recentemente, um novo método de redução de ruído, denominado C2F, foi desenvolvido especificamente para implantes cocleares (IC). Esse método resulta em maior inteligibilidade da fala em usuários de IC, em comparação com o filtro de Wiener (WF), quando os coeficientes são calculados de forma ideal. Em aplicações reais, os coeficientes são obtidos a partir de estimativas da razão sinal-ruído (SNR) e, portanto, acarretam perda do desempenho ótimo. Neste trabalho é realizada uma análise preliminar da robustez do método C2F a erros de estimação da SNR. Simulações numéricas corroboram a análise teórica mostrando que, apesar de apresentar diminuição de desempenho na redução de ruído, o C2F é menos sensível que o WF a erros de estimação da SNR.

Download
Estudo de Perdas para Codificação de Geometria de Nuvens de Pontos por Decomposição Diádica
Davi R. Freitas, Eduardo Peixoto, Ricardo L de Queiroz, Edil Medeiros

DOI: 10.14209/SBRT.2020.1570658416
Keywords: Nuvens de pontos Codificação com perdas Compressão de geometria
Abstract
This paper proposes a study about the introduction of lossy techniques over an intra-frame coder of the geometry information of voxelized point clouds (PC). Using an alternative representation approach, this method represents the PC as an array of binary images. Moreover, we sought to assess the quality of the decoded PC with different reconstruction techniques for a given bit rate. We observed that the use of the lossy proposed method alone wasn't enough to reach bitrates values below 0.2 bpov, which suggests the combination of this method with other lossy solutions in future studies.

Download
Detecção de tráfego anômalo de rede utilizando clusterização em Big Data
Mateus Rocha, Daniel G Silva

DOI: 10.14209/SBRT.2020.1570658430
Keywords: Big Data clustering network intrusion detection
Abstract
Nowadays, the sheer amount of information sent through the Internet enables the adoption of Big Data and machine learning frameworks in order to detect network anomalies. However, there are two key challenges: the processing latency due to huge amounts of data and the reduced flexibility that supervised learning paradigm may cause. In this paper, we propose a Big Data framework that uses unsupervised learning for near real time intrusion detection, which is also capable of periodically retrain the generated models in order to track the network dynamics. The framework is succesfully tested with a well-known dataset and with real network traffic from a reverse proxy server.

Download
  Página de 26