GNU Radio's DPD Package
|
LMS based Algorithm implemented to estimate the coefficients of the behaviorial model (GMP) of the Power Amplifier and thus predistorter taps. More...
#include <LMS_postdistorter.h>
Public Types | |
typedef boost::shared_ptr< LMS_postdistorter > | sptr |
Static Public Member Functions | |
static sptr | make (const std::vector< int > &dpd_params, int iter_limit, std::string method) |
Make LMS_postdistorter. More... | |
LMS based Algorithm implemented to estimate the coefficients of the behaviorial model (GMP) of the Power Amplifier and thus predistorter taps.
It estimates the coefficients based on the PA_output and PA_input. It uses the LMS (Least Mean Squares) Algorithm with two methods options, namely, Newton based and EMA based methods.
It has two input ports, one for the PA_output (gain phase calibrated) and other for the PA_input (or predistorter output).
The coefficients estimated are passed as messages through message output port 'taps'. Both Input ports are only for Complex Data Type.
typedef boost::shared_ptr<LMS_postdistorter> gr::dpd::LMS_postdistorter::sptr |
|
static |
Make LMS_postdistorter.
dpd_params | The (K_a, L_a, K_b, L_b, M_b) int_vector denoting the GMP model parameters used for predistorter 'taps' estimation. Total No. of coefficients = ((K_a * L_a) + (K_b * M_b * L_b)) |
iter_limit | Iteration limit or Max. number of iterations of training to be performed for predistorter DPD coefficients estimation. |
method | Method of LMS algorithm used for coefficients estimation, i.e., Newton or EMA based method. |