Keith Finnerty

Behavioral Modeling of Outphasing Amplification Systems

This paper presents a comparison of existing and novel behavioral models targeted at the outphasing power amplifier (PA) architecture. A comprehensive comparison of ten modeling strategies is presented in the results. Novel techniques for outphasing PAs, such as vector switched and dual path time series, are also presented for the first time. Investigation of such techniques was driven by the analysis of outphasing operation at minimum output powers, demonstrating the generation of frequency-dependent amplitude and phase deviations, which can be difficult to characterize. The increased robustness was achieved at the cost of additional complexity; for practical implementation, time series coefficient reduction techniques were also evaluated. The results of all modeling approaches are experimentally validated for the wideband operation of an NXP 19-W GaN digital outphasing amplifier module. Considering computational complexity and accuracy for system-level modeling across all presented options, a subset of existing and new models is identified as best suited for modeling outphasing PAs.

Linearization of an Outphasing Amplifier for Wide-band Multi-carrier Signals

A novel method using self-organizing maps to determine the optimum ranges to partition the vector switched Volterra series is presented. Analyzing characteristics of the input signal it is possible to separate different regions of operation. Partitioning of the correction algorithm provides the ability to target regions of different non-linear behavior with separate coefficients. This work is validated with experimentally measured results and shows an adjacent channel power ratio (ACPR) of -49/-50 dBc achieved for a four carrier Wideband Code-Division Multiple Access (WCDMA) signal using a 19 watt digital outphasing amplifier from NXP.

Behavioral modeling approach for array of amplifiers in active antenna array system

Memory polynomial behavioral models are widely used to model nonlinear distortion mechanisms in a single RF power amplifier. However, when a number of power amplifiers are used in parallel, as in the case of active antenna arrays, the total number of parameters will increase significantly. Hence, new strategies are required to model and linearize these systems efficiently. In this paper a novel efficient modeling approach for these array systems is presented and experimentally verified by applying the method to the measured data of an array of amplifiers. The experimental results prove the efficiency of the approach.

Digital quadrature mixing of lowpass sigma-delta modulators for switch-mode power amplifiers

In this paper a phase compensation technique for the digital up conversion of a quadrature signal for amplification with switch mode power amplifiers is proposed. When a digital signal generator is used to generate the complex envelope signal care must be taken to compensate for the phase skew between the two paths. If phase compensation is not implemented an image caused by up converting the complex envelope of the modulation signal is created. By compensating for phase skew between the I and Q signal paths it is possible to remove this image signal and enable the transmission of multi carrier signals. As a direct result of this technique there is a reduction in the filtering effort at the output of the power amplifier to meet spectral mask requirements.

A low complexity pre-distortion scheme for power amplifier linearization in wideband applications

Summary In this paper, a nonlinear autoregressive with exogenous inputs (NARX) digital pre-distortion scheme to linearize power amplifiers is proposed. The proposed NARX digital pre-distortion gives better accuracy and spectral leakage suppression compared with other commonly used Volterra-based techniques. The stability criterion of the NARX digital pre-distortion is derived from the frequency domain analysis. Simulation is carried out with a 20 MHz single carrier long-term evolution signal and two-carrier long-term evolution signal and instantaneous to average ratio of 6.2 dB at 0.01% complementary cumulative distribution function (CCDF). The results of simulation analysis show a slight improvement in adjacent channel leakage ratio performance with 30% reduction in the number of floating point operations compared with conventional pre-distortion techniques. Copyright

A Digital Pre-Distortion Based on Nonlinear Autoregressive With Exogenous Inputs

In this letter, a new pre-distortion technique for power amplifiers in wideband applications is proposed. The proposed pre-distortion technique is based on Nonlinear Autoregressive with Exogenous inputs (NARX). The forward path of the proposed predictive method is based on the memory polynomial. Experimental validation is carried out with 4 carrier WCDMA signal with 20MHz bandwidth and PAPR = 9.8 dB. The results show significant reduction in the number of coefficients with comparable performance in terms of adjacent channel leakage ratio (ACLR) and error vector magnitude (EVM) to Volterra series techniques.

Analytical investigation of thermal effects on outphasing power amplifiers

The outphasing amplifier has alternative linearity requirements when compared with traditional amplifier technology. As a relatively new architecture for mobile communications the physical and electrical causes of nonlinearity have not yet been fully identified. In this work the effect of thermal variations on the characteristic operation of the outphasing amplifier architecture is investigated. More specifically the impact that gain and phase variations which result from a change in device temperature affect the characteristic performance of the outphasing amplifier. Amplifier biasing circuits commonly account for changes in temperature, a process which will result in a shift in characteristic performance. The result demonstrates a shift in characteristic outphasing angle of as much as 2.4 degrees and a reduction in dynamic range greater than 20 dB.

Estimation of Sparse Memory Taps for RF Power Amplifier Behavioral Models

When a larger than required dimension such as memory depth or order of nonlinearity, is specified during behavioral model extraction, redundant terms can be calculated when determining the weights of the model. Extraction of a behavioral model can therefore benefit from a priori knowledge of the system to be modeled. Conversely if there is a limitation in the hardware required to calculate model outputs a limit can be set for the maximum number of weights to be used. In this letter, an approach is proposed which allows the input delay vector to be reduced to a sparse vector including the delayed samples which are most important in the construction of the power amplifier model. Simulations of behavioral models for experimentally measured data of two different PAs demonstrates the sparse models extracted in this way are as accurate as a full model but have a more compact and as a result more computationally efficient structure.