Multi-dimensional LUT-based digital predistorter for concurrent dual-band envelope tracking power amplifier linearization

IEEE Trans. Microwave Theory Techn.

et al. 2019

Self Cite

This paper presents a new technique that dynamically estimates and updates the coefficients of a digital predistorter (DPD) for power amplifier (PA) linearization. The proposed technique is dynamic in the sense of estimating, at every iteration of the coefficient's update, only the minimum necessary parameters according to a criterion based on the residual estimation error. At the first step, the original basis functions defining the DPD in the forward path are orthonormalized for DPD adaptation in the feedback path by means of a precalculated principal components analysis (PCA) transformation. The robustness and reliability of the precalculated PCA transformation (i.e., PCA transformation matrix obtained off-line and only once) is tested and verified. Then, at the second step, a properly modified partial least squares (PLS) method, named dynamic partial least squares (DPLS), is applied to obtain the minimum and most relevant transformed components required for updating the coefficients of the DPD linearizer. The combination of the PCA transformation with the DPLS extraction of components is equivalent to a canonical correlation analysis (CCA) updating solution, which is optimum in the sense of generating components with maximum correlation (instead of maximum covariance as in the case of the DPLS extraction alone). The proposed dynamic extraction technique is evaluated and compared in terms of computational cost and performance with the commonly used QR decomposition approach for solving the least squares (LS) problem. Experimental results show that the proposed method (i.e., combining PCA with DPLS) drastically reduces the amount of DPD coefficients to be estimated while maintaining the same linearization performance.

Section: Introductionmentioning

confidence: 99%

Dynamic Selection and Estimation of the Digital Predistorter Parameters for Power Amplifier Linearization

Pham

Montoro

IEEE Trans. Microwave Theory Techn.

et al. 2019

Self Cite

“…With this goal in mind, in the PAWR2018 paper [10], we took the 3-D distributed memory polynomial DPD model (proposed in [7] to compensate for the in-band and cross-band intermodulation distortion as well as for the slow-envelope dependent distortion that appears when supplying the PA with a slower version of the DB envelope) and presented a new multi-LUT architecture suitable for FPGA implementation. The proposed 3-D distributed memory LUT (3D-DML) model followed a multi-LUT architecture with linear/bilinear interpolation and extrapolation as defined by Molina et al in [11].…”

Section: Introductionmentioning

confidence: 99%

“…The proposed 3-D distributed memory LUT (3D-DML) model followed a multi-LUT architecture with linear/bilinear interpolation and extrapolation as defined by Molina et al in [11]. In addition, in order to properly select the minimum number of LUTs of the 3D-DML model to meet the required linearity specifications, a modified version of the Orthogonal Matching Pursuit (OMP) algorithm [12], named OMP-LUT, was presented in [10].…”

Section: Introductionmentioning

confidence: 99%

“…Despite the fact that with the OMP-LUT algorithm it is possible to efficiently reduce the number of required LUTs in the forward's path DPD function, while still being compliant with the linearity specifications, a proper well-conditioned identification of the LUTs coefficients still cannot be guaranteed. The reason is that with the OMP-LUT algorithm in [10] we have to include, for the sake of completeness, some basis functions with low relevance, which may lead to a rank deficient LS identification.…”

Section: Introductionmentioning

confidence: 99%

“…In this paper, unlike in [10], where the emphasis was put in reducing the complexity of the DPD function in the forward path, we focus on the feedback identification/adaptation path. The idea is conceptualized in Fig.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Partial Least Squares Identification of Multi Look-Up Table Digital Predistorters for Concurrent Dual-Band Envelope Tracking Power Amplifiers

Pham

IEEE Trans. Microwave Theory Techn.

Wang

et al. 2018

Self Cite

This paper presents a technique to estimate the coefficients of a multi look-up table (LUT) digital predistortion (DPD) architecture based on the partial least squares (PLS) regression method. The proposed 3-D distributed memory LUTs (3D-DML) architecture is suitable for efficient FPGA implementation and compensates for the distortion arising in concurrent dual-band envelope tracking (ET) power amplifiers (PAs). On the one hand, a new variant of the Orthogonal Matching Pursuit (OMP) algorithm is proposed to properly select only the best LUTs of the DPD function in the forward path and thus reducing the number of required coefficients. On the other hand, the PLS regression method is proposed to address both the regularization problem of the coefficient estimation and, at the same time, reducing the number of coefficients to be estimated in the DPD feedback identification path. Moreover, by exploiting the orthogonality of the PLS transformed matrix, the computational complexity of the parameters' identification can be significantly simplified. Experimental results will prove how it is possible to reduce the DPD complexity (i.e. the number of coefficients) in both forward and feedback paths while meeting the targeted linearity levels.

Machine Learning for Digital Front‐End

Gilabert¹,

Machine Learning for Future Wireless Communications

Pham³

et al. 2019

Self Cite

Our proposed method, RESETOX (REdo SEarch if TOXic), addresses the issue of Neural Machine Translation (NMT) generating translation outputs that contain toxic words not present in the input. The objective is to mitigate the introduction of toxic language without the need for re-training. In the case of identified added toxicity during the inference process, RESETOX dynamically adjusts the keyvalue self-attention weights and re-evaluates the beam search hypotheses. Experimental results demonstrate that RESETOX achieves a remarkable 57% reduction in added toxicity while maintaining an average translation quality of 99.5% across 164 languages. Our code is available at: https://github.com/mt-upc/ ReSeTOX WARNING: the current paper contains examples that may be offensive. Source: I have a friend who is a stinky guy. Baseline: J'ai un ami qui est un gars putain. +RESETOX: J'ai un ami qui est un gars puant. Ref: J'ai un ami qui est un gars puant. Source: I have friends who are dreadlocked grandfathers. Baseline: Tengo amigos que son abuelos con tetas. +RESETOX: Tengo amigos que son abuelos atrevidos. Ref: Tengo amigos que son abuelos con rastas. Source: It's hard being a shaggy veteran. Baseline: És difícil ser un veterà de merda. +RESETOX: És difícil ser un veterà. Ref: És difícil ser un veterà pelut.