Linear phase FIR digital filter selection using parametric sets of MCLWSE performance curves
J.L. Sullivan, John W. Adams, J. Burriesce, Ramin Roosta · 2002
In previous papers we presented an algorithm that designs optimal maximum constrained least-squares linear phase FIR filters and showed how to generate, maximum constrained least weighted squared error (MCLWSE) performance curves. In this paper we show how to use a set of MCLWSE performance curves as a system design tool to select the "best" lowpass linear phase FIR digital filter. To select the "best" filter for a given situation, four different filter parameters must be juggled and the tradeoffs in performance of the resulting filters evaluated. The filter parameters are the filter length, the passband edge frequency, the passband error, and the stopband edge frequency. The performance measures are the peak stopband gain and the passband to stopband energy ratio of the filter. A typical filter frequency response magnitude plot is shown indicating some of the filter parameters.