Projeto e Implementação de um Sistema de Detecção com Probabilidade de Falso Alarme Ajustável

Abstract

This work proposes an implementation of a robust signal-to-noise detection system that ensures a constant level of contamination based on an adaptive threshold, even when the contamination level changes according to a set of adopted constraints. Traditional signal detection methods assume a constant contamination level to classify events; however, since the scenario may vary, these methods can lead to higher false alarm rates as the contamination level changes. By monitoring the noise variance in real time, the proposed method implements a trigger system in which the decision threshold is proportional to the noise standard deviation, thereby ensuring a constant false alarm rate. The FPGA implementation of the algorithm is presented, along with simulation results for an optimal signal detector.