Iterative Learning Control for Signal Separation in Dual-RF Input Doherty Transmitter

Signal separation function is one of the key factors affecting the efficiency and linearity of the dual-RF input doherty transmitter (DIDT). In order to improve the linearity of DIDT, an optimization method of signal separation function based on iterative learning control (ILC) is proposed in this paper. Moreover, efficiency dictionaries which derived from the static characterization results are employed to ensure that the drain efficiency of final stage power amplifier does not degrade during iteration. The experimental results show that, for the 20 MHz test signal, the proposed method can improve adjacent channel power ratio (ACPR) of the output signal from -24.2 dBc to -49.6 dBc, meanwhile the average output power is increased form 37.8 dBm to 38.4 dBm, and the corresponding drain efficiency is 46.2% and 49.5% respectively. In this situation, auxiliary linearization techniques, such as DPD, can be omitted from the DIDT.