Choosing a FIR filter order for the linearization of an analog LPF phase response

In digital processing of wideband signals, a high-order low-pass filter (LPF) is usually used for prefiltering an analog signal. This allows to use sampling frequency 3-4 times higher than the upper frequency of the signal spectrum. The phase response of a high order LPF has an essential non-linearity within its bandwidth, which appears as an irregularity of group-delay characteristic (GDC). Along with the irregularity of an amplitude response, the GDC irregularity leads to a significant change in the waveform, which in many cases is unacceptable. In this report, a method of reducing the GD. irregularity is proposed. It is implemented by using a digital FIR filter applied after the ADC in a digital processing device. The approach is the FIR filter implementation by using discrete samples of analog filter impulse response, taken in the reverse sequence. With this approach, the reduction of GD. irregularity depends on the various factors such as an analog LPF order, impulse response sampling frequency, and a number of finite impulse response samples used. The impact of these factors combination on linearization efficiency was not considered before. The report presents the results of a detailed study of how the above factors impact on GD. irregularity reduction index and shows the possibility of significant reducing the GD. irregularity of high order filters with a low order of the FIR filter. Since the amplitude frequency response of linearization filter practically agrees with the one of the analog LPF, the resulting response becomes more rectangular.