In this paper, we propose a new closed-loop learning architecture for digital predistortion (DPD) with piecewise (PW) memory polynomial models. The technique is targeted specifically for power amplifiers (PAs) that exhibit strong nonlinear behavior and nonlinear memory effects, such as those implemented with gallium nitride (GaN) technology. The learning algorithm is based on a computationally simple decorrelating learning rule, which is applied on each PW polynomial model separately. Measurements with LTE-A signals on a basestation GaN PA show that the proposed technique clearly outperforms the reference closedloop memory polynomial DPD, in terms of reducing the adjacent channel emissions.
|International Symposium on Wireless Communication Systems (ISWCS)
|INTERNATIONAL SYMPOSIUM ON WIRELESS COMMUNICATION SYSTEMS
|1/01/00 → …