Background: Power amplifiers (PAs) have nonlinearity, which can be modeled with different nonlinear functions. Lets assume that the PA output at time 't' is a function of inputs at time 't', 't-1', ..., 't-M', where 'M'is the memory of PA. In this script, we use random inputs and feed them to a real PA to measure the PA outputs. We use these inputs and outputs to model the PA nonlinearities with a neural network.
To run the code, take the following steps: 1- Run MATLAB code to generate the required .mat files or use already generated (uploaded) .mat files 2- Run Python code to model the Power Amplifier nonlinearities
Notes: 1- "weblab" is used to access output data from a real power amplifier. Random data input is fed into the PA, and the output is read from "weblab". Then, a fully connected neural network is used to model the PA behaviour. 2- To generate random constellation we used "randconst" function. You should either write your own code to generate random constellaion or ask "RFWebLab@dpdcompetition.com" to access the code.
For more info about "weblab" please see http://www.dpdcompetition.com/