Uma Nova Função de Aproximação Least Square com Zeros Imaginários Obtidos por Otimização

In this paper, a special class of approximation functions is studied using the least-squares method, which yields a minimal of power loss in passband. This method allows the choice of some proprieties, such as either monotonic or nonmonotonic, weighting of some passband frequencies, band choice to be minimized, and inclusion of imaginary zeros. In addition, a new approximation function (SN) is presented, which uses a least-squares characteristic function, applying again the same method to optimize the bandstop. This optimization provides imaginary zeros, which makes the obtained function nearest of a brick-wall filter.