Search

ABSTRACT This paper addresses the problem of model size estimation in power amplifiers behavioural modeling and predistortion applications using multi-basis functions memory polynomials. A low complexity sequential approach is proposed along with a suitable model. The proposed model and size estimation approach are experimentally assessed for the modeling and predistortion of an LTE envelope-tracking power amplifier. © 2016 Wiley Periodicals, Inc. Microwave Opt Technol Lett 58:726-732, 2016 [ABSTRACT FROM AUTHOR]

The demand for high data rates and broadband wireless access necessitates the development of radio systems that use multiband signals. With the deployment of these systems, the radio-frequency front ends and power amplifiers (PAs) used exhibit many nonlinear effects leading to signal distortions. The use of a linearization technique is necessary to enhance the linearity–efficiency tradeoff. This brief explores new schemes suitable for the modeling and linearization of transmitters suffering from hardware imperfections when driven with dual-band signals. An enhanced and accurate 2-D Cartesian memory polynomial model (2DECMPM) is proposed to compensate for hardware impairments such as the in-phase and quadrature-phase (I–Q) imbalance and dc offset in addition to nonlinear distortion. From an accuracy perspective, the proposed model offers better compensation performance, mainly due to the use of additional cross-terms between I and Q signals compared to the 2-D joint memory polynomial model (2DJMPM). Experimental results show that the proposed 2DECMPM-based predistorter enabled superior PA linearization and I–Q imbalance compensation with a comparable complexity with the state-of-the-art 2DJMPM model. [ABSTRACT FROM PUBLISHER]

This paper presents a method for pruning power amplifier (PA) behavioral models and digital predistorters based on compressed sampling theory. Using this method, the number of coefficients required by behavioral models and predistorters can be significantly reduced while achieving comparable performance in terms of both modeling accuracy and suppressing distortions. Hardware measurements obtained for a Doherty PA driven by a five-carrier 100-MHz wide long-term evolution-advanced signal demonstrate the capability of the proposed method to linearize a highly nonlinear PA prototype using a minimal number of coefficients, revealing the attractive properties of the proposed method and its desirable performance. Using the proposed technique, predistorters achieving similar linearization performance while requiring significantly less coefficients than the traditional models were demonstrated. [ABSTRACT FROM AUTHOR]