A Weighted Linearization Method for Highly RF-PA Nonlinear Behavior Based on the Compression Region Identification

Full-duplex (FD) communication systems allow for increased spectral efficiency but require effective self-interference cancellation (SIC) techniques to enable the proper reception of the signal of interest. The underlying idea of digital SIC is to estimate the self-interference (SI) channel based on the received signal and the known transmitted waveform. This is a challenging task since the SI channel involves, especially for mass-market FD transceivers, many nonlinear distortions produced by the impairments of the analog components from the receiving and transmitting chains. Hence, this paper first analyzes the power of the SI components under practical conditions and focuses on the most significant one, which is proven to be produced by the I/Q mixer imbalance. Then, a widely-linear digital SIC approach is adopted, which simultaneously deals with the direct SI and its image component caused by the I/Q imbalance. Finally, the performances achieved by linear and widely-linear SIC approaches are evaluated and compared using an experimental FD platform relying on software-defined radio technology and GNU Radio. Moreover, the considered experimental framework allows us to set different image rejection ratios for the transmission path I/Q mixer and to study its influence on the SIC capability of the discussed approaches.

Section: Power Amplifier Nonlinearitymentioning

confidence: 99%

“…The PA nonlinear distortion has been widely studied in simulated and experimental approaches, and different estimations have been proposed [14][15][16][17]. Additionally, the I/Q imbalance has been identified as another significant distortion source engendering a strong SI image component in the SI channel of FD transceivers.…”

Section: Introductionmentioning

confidence: 99%

Widely-Linear Digital Self-Interference Cancellation in Full-Duplex USRP Transceiver

Despina-Stoian

Youssef

Digulescu

et al. 2022

“…The model FV(13,3) contains a raw stock of 248 regressors. The evolution of the NMSE and the BIC as the greedy search includes more active regressors (coefficients) is shown in Figure 1, indicating a model reduced to eight active regressors, denoted as s-FV (13,3), with an NMSE of −55.4 dB. The normalized magnitude of the corresponding estimated coefficients, labeled with the associated regressors, are displayed in the upper plot of Figure 2.…”

Section: Algorithm 1: Upgrading An Incomplete Modelmentioning

confidence: 99%

“…Regrettably, the size of its regressor set can be unsuitably large, and ad hoc models with a reduced set of regressors were proposed. For example, the memory polynomial (MP) model and its modifications [ 12 , 13 ], together with the generalized memory polynomial (GMP) [ 2 ] model, have demonstrated satisfactory performance in the design of a DPD. Another approach based on available information at the circuit level has also made possible the deduction of a reduced-order structure that contains the GMP lagging envelope terms with even-order envelope powers as a particular regressor type [ 14 ].…”

Section: Introductionmentioning

confidence: 99%

Upgrading Behavioral Models for the Design of Digital Predistorters

Crespo‐Cadenas

Madero‐Ayora

Becerra

2021

This work presents a strategy to upgrade models for power amplifier (PA) behavioral modeling and digital predistortion (DPD). These incomplete structures are the consequence of nonlinear order and memory depth model truncation with the purpose of reducing the demand of the limited computational resources available in standard processors. On the other hand, the alternative use of model structures pruned a priori does not guarantee that every significant term is included. To improve the limited performance of an incomplete model, a general procedure to augment its structure by incorporating significant terms is demonstrated. The sparse nature of the problem allows a successive search incorporating additional terms with higher nonlinear order and memory depth. This approach is investigated in the modeling and linearization of a commercial class AB PA operating at a compression point of about 6 dB, and a class J PA operating near saturation. Results highlight the capabilities of this upgrading procedure in the improvement of linearization capabilities of DPDs.

“…In addition, Radio-frequency/Microwave transmission circuits with wide dynamic bandwidths represent a challenge for modeling and DPD linearization techniques. Behavioral modeling based on higher-order polynomial systems for long-term and short-term induced memory represents an efficient methodology for improving spectral efficiency [ 1 , 2 ]. The selection of the primary method for the modeling and linearization stages is essential if it is required to use multivariable data, gain precision, and not increase the involved coefficients during hardware implementation [ 3 ].…”

Section: Introductionmentioning

confidence: 99%

A Comparison of Surrogate Behavioral Models for Power Amplifier Linearization under High Sparse Data

Galaviz-Aguilar

Vargas-Rosales²,

Cárdenas-Valdez³

et al. 2022

Self Cite

A good approximation to power amplifier (PA) behavioral modeling requires precise baseband models to mitigate nonlinearities. Since digital predistortion (DPD) is used to provide the PA linearization, a framework is necessary to validate the modeling figures of merit support under signal conditioning and transmission restrictions. A field-programmable gate array (FPGA)-based testbed is developed to measure the wide-band PA behavior using a single-carrier 64-quadrature amplitude modulation (QAM) multiplexed by orthogonal frequency-division multiplexing (OFDM) based on long-term evolution (LTE) as a stimulus, with different bandwidths signals. In the search to provide a heuristic target approach modeling, this paper introduces a feature extraction concept to find an appropriate complexity solution considering the high sparse data issue in amplitude to amplitude (AM-AM) and amplitude to phase AM-PM models extraction, whose penalties are associated with overfitting and hardware complexity in resulting functions. Thus, experimental results highlight the model performance for a high sparse data regime and are compared with a regression tree (RT), random forest (RF), and cubic-spline (CS) model accuracy capabilities for the signal conditioning to show a reliable validation, low-complexity, according to the peak-to-average power ratio (PAPR), complementary cumulative distribution function (CCDF), coefficients extraction, normalized mean square error (NMSE), and execution time figures of merit. The presented models provide a comparison with original data that aid to compare the dimension and robustness for each surrogate model where (i) machine learning (ML)-based and (ii) CS interpolate-based where high sparse data are present, NMSE between the CS interpolated based are also compared to demonstrate the efficacy in the prediction methods with lower convergence times and complexities.