Análise das Distribuições SαS e αSG na Modelagem do Ruído de Camarão-de-Estalo

Abstract

Resumo-Sistemas de comunicação subaquáticos por ondas acústicas estão sujeitos à degradação de desempenho causada pelo ruído impulsivo gerado pelo crustáceo camarão-de-estalo. Em análises matemáticas ou simulações, esse ruído é comumente modelado por meio de amostras temporalmente descorrelacionadas seguindo a distribuição alfa-estável simétrica (symmetric alpha-stable, SαS). Neste artigo analisa-se a adequação dessa distribuição em contraste com a distribuição alfa-sub-Gaussiana (alpha-sub-Gaussian, αSG) com correlação temporal não nula, no contexto do sensoriamento espectral subaquático.Palavras-Chave-Camarão-de-estalo, comunicação subaquática, ruído impulsivo, sensoriamento espectral.

Direct-Conversion Spectrum Sensor Impaired by Symmetric α-Stable and α-Sub-Gaussian Noises

“…Conventional and DCR-based receiver models are used to compare the performances of the energy detector (ED), the absolute value cumulating (AVC) detector, the Gerschgorin radii and centers ratio (GRCR) detector, the Gini index detector (GID), the Pietra-Ricci index detector (PRIDe) and the locally most powerful invariant test (LMPIT) detector. This work merges and extends the research reported in [17] and [27]. The merged part refers to the assessment of the conventional and the DCR-based receiver models under SαS and αSG noise.…”

“…References [20]- [25] are cited as examples. In [17], the adequacy of the SαS distribution is analyzed in contrast to the αSG distribution with non-zero temporal correlation, in the context of underwater spectrum sensing subjected to snapping shrimp noise.…”

“…The merged part refers to the assessment of the conventional and the DCR-based receiver models under SαS and αSG noise. The extended part refers to the addition of a large number of new results in which the spectrum sensing performance is measured as a function of the main system parameters, instead of using receiver operating characteristic (ROC) curves, for all detectors considered in [27] and [17], and for an additional detector, the AVC, claimed to be suitable for the impulsive noise scenario.…”

“…Commonly, in mathematical analysis and simulations, the snapping shrimp noise is modeled through temporally uncorrelated samples, following the symmetric alpha-stable (SαS) distribution [9], [14]. However, it is suspected that the SαS distribution is not the most suitable choice, which has been also verified, for instance, in [15]- [17]. This is owed to the fact that the snapping shrimp noise samples exhibit a temporal correlation that is not present in SαS samples.…”