Digital Pre-Distorter Based On A Box Oriented Memory Polynomial Model and Optimized By Tabu Search Algorithm for Wimax Radio Frequency Power Amplifier

Digital predistortion (DPD) technique is widely used to linearize a radio frequency power amplifier (PA). It is the fastest and cost effective digital processing technique to moderate the distortions in PA caused by nonlinearity and memory effects. Digital predistorter is inverse of the PA, so the design of a digital predistorter requires the accurate identification of PA’s behavior. This paper presents a box oriented memory polynomial DPD technique based on an indirect learning approach. The adjacent channel leakage ratio (ACLR), error vector magnitude (EVM) and number coefficients of the PA and digital predistorter are evaluated of the proposed model. The coefficients of PA and DPD are optimized iteratively in order to minimize the output PSD around the pre specified frequency. In the proposed work metaheuristic optimization algorithm called tabu search algorithm (TSA) is used to optimize the coefficients of PA and digital predistorter. This approach reduced the complexity and cost of DPD technique implementation. The TSA produces hopeful results for Worldwide Interoperability of Microwave Access (WiMAX) signal. KeywordsPower Amplifier, Memory Polynomial, Digital Predistorter, tabu search algorithm, radio frequency, WiMAX, Power Spectral Density etc.