The Design And Implementation Of Broadband Digital Pre-distortion System

With the development of wireless communication technology, the capacity of communication system will be larger and larger, and the requirement for the quality of communication will also be higher and higher. 3G, 4G and 5G which is being studied, have higher requirement for spectrum utilization, so all of them usually use high order modulation techniques. Although these modulation techniques have high spectrum efficiency, they can be affected by the nonlinear of the system easily. PA is a key part of radio transmitter and because of its inherent nonlinear and the characteristics of the latest modulation methods, nonlinear distortion of the output signal can occur easily; and this leads to the decrease of the quality of communication in the communication system. PA linearization techniques have existed about one hundred years ago. In varieties of PA linearization techniques, digital pre-distortion technology is the most promising linearization technology. Because digital pre-distortion can be implemented simply, and has relatively low cost, high flexibility, wide bandwidth, and good linear effect, it has become the most popular pre-distortion technique.At present, digital pre-distortion devices are mainly used in the narrow band communication systems, which can't meet the development trend of broadband communication. Memory distortion in broadband communication system cannot be neglected, so this paper designs and implements a broadband digital pre-distortion system to solve this problem.This scheme selects a linearization method which is based on look-up table. Because look-up table does not consider memory, this paper uses a FIR filter to correct PA's memory distortion, and use the updating algorithm to fulfill adaptive updating. Simultaneously, this paper completes the system simulation by using Matlab and verifies that this scheme is feasible. On this basis, this paper designs a pre-distortion circuit board, and implements the lookup table algorithm and gets the filter tap coefficients, and completes adaptive updating by putting the data into the ASIC module of the FPGA. Finally, in the environment of Quartus II 10.1 and Nios II 10.1 Software Build Tools for Eclipse Software, this paper tests the function and performance of this digital pre-distortion system by using spectrum analyzer, and the results achieve the expected targets.