Search

In this paper, we present an EVM-BER conversion method applicable to PAM/QAM modulations. Using SNR as an indicator of the transmission environment, we analyze the EVM and BER of the modulations. EVM-BER conversion is performed by determining the SNR value that corresponds to a given value of EVM using the bisection method, and then sequentially obtaining the BER value for this SNR. By measuring the output of power amplifiers equipped with a DPD using QAM-modulated signals as the input, the accuracy of the proposed conversion method in predicting the BER was verified by comparing it with the existing method. [ABSTRACT FROM AUTHOR]

This paper presents a novel magnitude‐selective affine (MSA) based digital predistortion (DPD) method for the performance investigation of multiband 5G new radio (NR) based analog radio over fiber link. The proposed MSA‐DPD method is derived from the canonical piecewise linear (CPWL) based model by employing MSA functions, which can result in reduction of the number of multiplication operations and the model complexity. The 5G NR standard at 20 GHz with 50 MHz bandwidth and flexible‐waveform signal at 3 GHz with 20 MHz bandwidth is used. A dual drive Mach Zehnder modulator having two distinct RF signals modulates a 1310 nm optical carrier using distributed feedback laser for 22 km of standard single mode fiber. The proposed MSA‐DPD method is compared with the CPWL and generalized memory polynomial method. The experimental results are presented in terms of adjacent channel power ratio, error vector magnitude, number of estimated coefficients and multiplications suggesting that MSA‐DPD method achieves a better performance as compared to CPWL and GMP models with much lesser complexity meeting the 3GPP Release 17 requirements. [ABSTRACT FROM AUTHOR]

This paper proposes an efficient neural-network-based digital predistortion (DPD), named as envelope time-delay neural network (ETDNN) DPD. The method complies with the physical characteristics of radio-frequency (RF) power amplifiers (PAs) and uses a more compact DPD model than the conventional neural-network-based DPD. Additionally, a structured pruning technique is presented and used to reduce the computational complexity. It is shown that the resulting ETDNN obtained after applying pruning becomes so sparse that its complexity is comparable to that of conventional DPDs such as memory polynomial(MP) and generalized memory polynomial (GMP), while the degradation in performance due to the pruning is negligible. In an experiment on a 3.5-GHz GaN Doherty power amplifier (PA), our method with the proposed pruning had only one-thirtieth the computational complexity of the conventional neural-network-based DPD for the same adjacent channel leakage ratio (ACLR). Moreover, compared with memory-polynomial-based digital predistortion, our method with the proposed pruning achieved a 7-dB improvement in ACLR, despite it having lower computational complexity.

This letter presents a novel neural network (NN) based digital predistortion (DPD) technique to obliterate the signal impairments and nonlinearities in radio over fiber (RoF) systems. DPD is generally performed with volterra based procedures that utilizes indirect learning architecture (ILA) that can become complex and expensive computationally. The proposed method using NNs evades issues associated with ILA and utilizes a NN to first model the RoF link and then training a NN based predistorter by backpropagating through the RoF NN model. Furthermore, the experimental evaluation is carried out for long term evolution 20‐MHz 256‐QAM modulation signal using 1310 nm distributed feedback laser, and standard single‐mode fiber to establish a comparison between NN based RoF link and volterra based generalized memory polynomial using ILA. The efficacy of the DPD is examined by reporting adjacent channel power ratio, mean square error, and error vector magnitude. The experimental findings imply that NN‐DPD convincingly learns the RoF nonlinearities which may not suit a volterra based model, and hence may offer a favorable trade‐off in terms of computational overhead and DPD performance. [ABSTRACT FROM AUTHOR]

This letter presents a comparative evaluation between three different behavioral models to perform digital predistortion (DPD) that enhances the linearity of radio‐over‐fiber (RoF)‐based front haul links for the mobile network. In particular, the intention is to jump out of the volterra box and propose models based on segmentation approach. Especially the decomposed vector rotation (DVR) model is compared to volterra polynomials such as memory and generalized memory polynomial (GMP) architectures. DPD is employed to RoF links that are based on distributed feedback laser emitting at 1310 nm, and standard single‐mode fiber for long‐term evolution 20‐MHz signal with 256‐QAM modulation format. The effectiveness of the digital predistortion methodology is investigated for varying input powers in terms of normalized mean square error, adjacent channel power ratio, and error vector magnitude. The experimental results demonstrate that DVR achieves elevated linearization when compared to memory polynomial and GMP models. [ABSTRACT FROM AUTHOR]

The generalized memory polynomial (GMP) model is one of the most commonly used models in digital predistortion to compensate for the nonlinearities and memory effects of power amplifiers (PAs). A very difficult, yet crucial, aspect of behavioral modeling is model dimensioning, i.e., determining nonlinear orders and memory depths. In this paper, we propose an algorithm based on hill-climbing (HC) heuristic to search for a GMP model structure, which provides the best tradeoff between modeling accuracy and complexity with different search criteria. Algorithm performance and convergence are evaluated for two different search criteria. Some optimization methods of the algorithms are then proposed to reduce the execution time by controlling its search path. The algorithm has been evaluated using real measurements from two different Doherty PAs. The solutions of the algorithm are proved robust in linearization of PA on the test bench. The results show that the proposed algorithm allows us to decrease the searching time by a factor typically greater than $10^4$ compared with exhaustive search. [ABSTRACT FROM AUTHOR]

A new learning-based structure is defined for the identification of nonlinear attributes of a baseband power amplifier (PA) and design of an adaptive digital pre-distorter for application in wireless channels. Analytically and experimentally, generalized memory polynomial (GMP) model is derived from the transmit–receive conditions and the learning network structure. The test data are a sequences of 16-quadrature amplitude modulation (QAM) and 64-QAM complex symbols. In simulation, unknown GMP models of different order are used as a baseband PA model for identification. Minicircuits ZHL-4240 radio frequency PA modules along with NI USRP 2920 units are employed for experimental verification of the concept. The learning aspect is governed by a specially designed artificial neural network structure called real valued focused time delay neural network. Simulation and empirical results agree with analytical derivation. [ABSTRACT FROM PUBLISHER]

Accurately modeling the memory effects is critical in modeling and nonlinearity pre-compensation for wideband radio frequency (RF) power amplifiers (PAs) with strong memory effects. In this paper, the memory effects are divided into two categories: lagging memory effect and leading memory effect, which are defined based on the input samples located after and before the current output samples, separately. Considering the above two types of memory effects, a generalized two-box cascaded (GTBC) nonlinear behavioral model for RF PAs with strong memory effects is proposed: one box for a static nonlinearity block that is used to include the strong static nonlinearities and the other box for a dynamic nonlinearity block which is used to predict the mild dynamic nonlinearities. The modeling accuracy of the proposed model, in terms of the normalized mean square error (NMSE) and the memory effect intensity (MEI), is compared to the generalized memory polynomial (GMP), the augmented Hammerstein (AH) and the enhanced Hammerstein (EH) models for a 450-470 MHz Doherty PA with a three-carrier WCMDA input signal and a 895-935 MHz inverse Class-F Doherty PA with a single-carrier FDD-LTE OFDM input signal. The validation demonstrates that the proposed model leads to high-accuracy with fewer modeling coefficients and floating point operations (FLOPs). [ABSTRACT FROM AUTHOR]