Hybrid Analog/Digital Linearization Based on Dual- Domain Decomposition of Nonlinearity

In this paper, the method of dual-domain decomposition of nonlinearity (DDDN) and dual-domain hybrid linearization (DDHL) of PA are proposed. The DDDN method decomposes the completed nonlinearity (CN) of power amplifier (PA) into static nonlinearity (SN) and dynamic nonlinearity(DN). The DDHL method uses a linearizer to compensate the SN, and a memory polynomial (MP) model for DN. Due to the components of DN is simplified and more similar at different working frequencies compared to CN, the number of parameters of the MP used in DDHL is reduced, and the broadband performance of linearization is improved compared to conventional MP model. A PA is tested with wideband signal at 3.4 and 3.5 GHz to verify the method of DDHL, the results show that the nonlinearity order is reduced from 8 (32 parameters) of MP-DPD to 5 (20 parameters) of DDHL when the ACPR improved from -30 dBc to -50 dBc.