Skip to main content

SOLUCIONES AVANZADAS DE PROCESADO DIGITAL DE SEÑAL PARA COMPONENTES Y SISTEMAS DE NUEVA RADIO 5G ENERGETICA Y COMPUTACIONALMENTE EFICIENTES

TEC2017-83343-C4-2-R

•
Funding agency name Agencia Estatal de Investigación
Funding agency acronym AEI
Program Programa Estatal de I+D+i Orientado a los Retos de la Sociedad
Call Proyectos I+D+i Retos de Investigación
Call year 2017
State Plan 2017-2020 Plan Estatal de Investigación Científica y Técnica y de Innovación
Beneficiary centre UNIVERSITAT POLITECNICA DE CATALUNYA

Resultados relacionados

Resultados totales (Incluyendo duplicados): 16
Encontrada(s) 1 página(s)

UPCommons. Portal del coneixement obert de la UPC
Publication Artículo científico (JournalArticle)

Nested fast adaptive cross approximation algorithm for solving electromagnetic scattering problems

  • Fang, Xiaoxing
  • Heldring, Alexander|||0000-0003-2011-2096
  • Rius Casals, Juan Manuel|||0000-0003-0606-5422
  • Chen, Xinlei
  • Cao, Qunsheng
In this paper, a Nested Fast Adaptive Cross Approximation (NFACA) algorithm is presented to accelerate the solution of electromagnetic scattering problems. It is based on the Fast Adaptive Cross Approximation (FACA) and the Nested Cross Approximation (NCA). In the NFACA, the FACA is applied instead of the ACA or the Chebyshev nodes, to enhance the efficiency. It is shown to have O(N) storage and computational complexity for electrically small problems. Different from the conventional NCA or the Fast Nested Cross Approximation (FNCA), the pivot-selection does not rely on a preliminary selection with Chebyshev nodes or ACA for each group. The NFACA permits the selection of the optimum pivots independently at each level to make the accuracy controllable. Numerical results of several objects are presented to demonstrate the performance of this method. Compared with the FACA and NCA in solving the electromagnetic scattering problems, the NFACA is more efficient both in computational cost and storage. Compared with the FNCA, the NFACA is both faster and more accurate., This work was supported in part by the National Nature Science Foundation of China under Grant 61871219, Grant 61771238, and Grant 61501227; in part by the Post-Graduate Research and Practice Innovation Program of Jiangsu Province under Grant KYCX18_0289; in part by the China Scholarship Council under Grant 201906830042; in part by the Spanish Ministerio de Ciencia e Innovacion (MICINN) under Project TEC2016-78028-C3-1-P, Project PID 2019-107885GB-C31, Project TEC2017-83343-C4-2-R, and Project MDM2016-0600; and in part by the Catalan Research Group 2017 SGR 219., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publication Artículo científico (JournalArticle)

Digital predistorters go multidimensional: DPD for concurrent multiband envelope tracking and outphasing power amplifiers

  • Gilabert Pinal, Pere Lluís|||0000-0001-6183-6977
  • Montoro López, Gabriel|||0000-0002-1328-4175
  • Vegas Garcia, Ángel
  • Ruiz Lavin, Maria de las Nieves
  • García García, José Ángel
© 2019 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works., Over at least the last two decades, digital predistortion (DPD) has become the most common and widespread solution to cope with the power amplifier's (PA's) inherent linearity-versus-efficiency tradeoff. When compared with other linearization techniques, such as Cartesian feedback or feedforward, DPD has proven able to adapt to the always-growing demands of technology: wider bandwidths, stringent spectrum masks, and reconfigurability. The principles of predistortion linearization (in its analog or digital forms) are straightforward, and the linearization subsystem precedes the PA (a nonlinear function in a digital signal processor in the case of DPD or nonlinear device in the case of analog predistortion and counteracts the nonlinear characteristic of the PA. Some excellent overviews on DPD can be found in [1]-[4]. Let us now look at the challenges that DPD linearization has faced and will continue to face in the near future with 5G new radio (5G-NR)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publication Artículo científico (JournalArticle)

Acceleration of Born series by change of variables

  • López Menchón, Héctor|||0000-0001-5776-1423
  • Rius Casals, Juan Manuel|||0000-0003-0606-5422
  • Heldring, Alexander|||0000-0003-2011-2096
  • Úbeda Farré, Eduard|||0000-0001-6759-0445
In this work, we propose a method to enhance the convergence of the Born series. The Born series is widely used in scattering theory, but its convergence is only guaranteed under certain restrictive conditions which limit the cases where this formulation can be applied. The proposed method, based on modifying the singularities of the resolvent operator by a change of variables, accelerates the convergence of the series or even achieves convergence in otherwise divergent cases. Due to its low computational cost and ease of implementation, the method preserves the advantages of the Born series while it extends its range of application. It can also be applied to a variety of methods relying on the same mathematical principles as the Born series. Numerical examples are provided to show that a significant improvement can be achieved with simple techniques and with a limited knowledge about the operator spectrum., This work was partly funded by the Ministerio de Ciencia e Innovacion (MICINN) under projects TEC2016-78028-C3-1-P, MDM2016-0600, TEC2017-83343-C4-2-R, TEC2017-84817-C2-2-R, PID 2019-107885GB-C31 and Catalan Research Group 2017 SGR 219; and by Secretary of Universities and Research of the Ministry of Business and Knowledge of the Generalitat of Catalonia and European Social Fund under scholarship 2019 FI B00788., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publication Contribución de congreso (ConferenceOutput)

A method for selecting online the coefficients to be updated in a DPD for PA linearization

  • Pham, Thi Quynh Anh|||0000-0001-5921-8184
  • Montoro López, Gabriel|||0000-0002-1328-4175
  • López Bueno, David|||0000-0002-9005-0532
  • Gilabert Pinal, Pere Lluís|||0000-0001-6183-6977
This paper presents a technique for selecting onlinethe coefficients to be updated in a digital predistorter (DPD)based on direct learning. The proposed method, which is basedon a combination of matching pursuit (MP) and least squares(LS) techniques (and is therefore named MP-LS method) allowsto improve the power amplifier (PA) linearization performanceof a fixed number of DPD coefficients, due to the fact that ateach DPD iteration the coefficients to be updated are properlychosen. The proposed technique is compared to a conventional LSestimation, and experimental results demonstrate that the MP-LSmethod can provide a performance improvement in relation toa DPD with fixed-preselected coefficients. The method could beespecially useful in DPD systems that have hardware restrictionsin the resources to be used by the update subsystem in thefeedback path. That is the case of DPDs based on FPGA devicesimplementing a QR algorithm in the programmable logic (PL)side, This work was partially supported by the Spanish Govern-ment (MICINN) and FEDER under project TEC2017-83343-C4-2-R, and by the Generalitat de Catalunya under grants 2017SGR 891 and 813., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publication Artículo científico (JournalArticle)

Improving the accuracy of the adaptive cross approximation with a convergence criterion based on random sampling

  • Heldring, Alexander|||0000-0003-2011-2096
  • Úbeda Farré, Eduard|||0000-0001-6759-0445
  • Rius Casals, Juan Manuel|||0000-0003-0606-5422
The accuracy of the Adaptive Cross Approximation (ACA) algorithm, a popular method for the compression of low rank matrix blocks in Method of Moment computations, is sometimes seriously compromised by unpredictable errors in the convergence criterion. This paper proposes an alternative criterion, based on global sampling of the error in the elements of the ACA compressed matrix. The sampling error depends on the size of the sample but also on the population distribution of the error, which makes it difficult to control the error independently of the underlying problem. However, as argued and demonstrated in the paper, the distribution of the error converges to the same unique probability distribution function for all low rank matrices. Complementing the sampling criterion with a simple mechanism to detect this convergence, we arrive at a criterion that controls the error irrespective of the underlying problem. As a practical example the RCS of a moderate size metallic ogive is computed to illustrate the merits of the proposed criterion. The proposed algorithm may also be useful in other methods that approximate low-rank matrices by interpolation of a reduced set of its elements., This work was partly funded by the Ministerio de Ciencia e Innovacion (MICINN) under projects TEC2016-78028-C3-1-P, TEC2017-84817-C2-2-R, TEC2017-83343-C4-2-R, PID 2019-107885GBC31, MDM2016-0600, and Catalan Research Group 2017 SGR 219., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publication Contribución de congreso (ConferenceOutput)

Dataset reduction for neural network based digital predistorters under strong nonlinearities

  • López Bueno, David|||0000-0002-9005-0532
  • Gilabert Pinal, Pere Lluís|||0000-0001-6183-6977
  • Montoro López, Gabriel|||0000-0002-1328-4175
The artificial neural networks (ANN) are gaining momentum in the digital predistorters (DPD) thanks to their inherently good approximation capabilities. Under strong or complex power amplifier nonlinearities, the size of the ANN can increase and lead to long training periods which are unaffordable in fast-changing waveform scenarios like those proposed for 5G or 6G. In this work we combine the orthogonal matching pursuit technique together with dataset length reduction methods, to significantly shorten the ANN DPD coefficients update time., This work was supported in part by the Spanish Government (Ministerio de Ciencia, Innovación y Universidades) and the Fondo Europeo de Desarrollo Regional (FEDER) under grants RTI2018-099841-B-I00 and TEC2017-83343-C4-2-R, and in part by the Generalitat de Catalunya under grants 2017 SGR 891 and 2017 SGR 813., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publication Artículo científico (JournalArticle)

Comparison of feature selection techniques for power amplifier behavioral modeling and digital predistortion linearization

  • Barry, Abdoul
  • Li, Wantao|||0000-0002-2634-6742
  • Becerra González, Juan Antonio
  • Gilabert Pinal, Pere Lluís|||0000-0001-6183-6977
The power amplifier (PA) is the most critical subsystem in terms of linearity and power efficiency. Digital predistortion (DPD) is commonly used to mitigate nonlinearities while the PA operates at levels close to saturation, where the device presents its highest power efficiency. Since the DPD is generally based on Volterra series models, its number of coefficients is high, producing ill-conditioned and over-fitted estimations. Recently, a plethora of techniques have been independently proposed for reducing their dimensionality. This paper is devoted to presenting a fair benchmark of the most relevant order reduction techniques present in the literature categorized by the following: (i) greedy pursuits, including Orthogonal Matching Pursuit (OMP), Doubly Orthogonal Matching Pursuit (DOMP), Subspace Pursuit (SP) and Random Forest (RF); (ii) regularization techniques, including ridge regression and least absolute shrinkage and selection operator (LASSO); (iii) heuristic local search methods, including hill climbing (HC) and dynamic model sizing (DMS); and (iv) global probabilistic optimization algorithms, including simulated annealing (SA), genetic algorithms (GA) and adaptive Lipschitz optimization (adaLIPO). The comparison is carried out with modeling and linearization performance and in terms of runtime. The results show that greedy pursuits, particularly the DOMP, provide the best trade-off between execution time and linearization robustness against dimensionality reduction.</jats:p>, This work was supported in part by the Spanish Government (Ministerio de Ciencia, Inno- vación y Universidades) and Fondo Europeo de Desarrollo Regional (FEDER) under Grants TEC2017- 83343-C4-2-R and PID2020-113832RB-C21 and in part by the Government of Catalonia and the European Social Fund under Grants 2017-SGR-813 and 2021-FI-B-137., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publication Artículo científico (JournalArticle)

Low-rank matrix factorization method for multiscale simulations: a review

  • Li, Mengmeng
  • Ding, Dazhi
  • Heldring, Alexander|||0000-0003-2011-2096
  • Hu, Jun
  • Chen, Rushan
  • Vecchi, Giuseppe
In this paper, a review of the low-rank factorization method is presented, with emphasis on their application to multiscale problems. Low-rank matrix factorization methods exploit the rankdeficient nature of coupling impedance matrix blocks between two separated groups. They are widely used, because they are purely algebraic and kernel free. To improve the computation precision and efficiency of low-rank based methods, the improved sampling technologies of adaptive cross approximation (ACA), post compression methods, and the nested low-rank factorizations are introduced. O(N) and O (NlogN) computation complexity of the nested equivalence source approximation can be achieved in low and high frequency regime, which is parallel to the multilevel fast multipole algorithm, N is the number of unknowns. Efficient direct solution and high efficiency preconditioning techniques can be achieved with the low-rank factorization matrices. The trade-off between computation efficiency and time are discussed with respect to the number of levels for low-rank factorizations., This work was supported in part by the Natural Science Foundation of China under Grant 61871222, Grant 62025109, Grant 61890541, Grant 61931021, and Grant 61721001; and in part by the Spanish Government under Project TEC 2017-83343-C4-2-R and Project MDM-2016-0600., Peer Reviewed
Recercat. Dipósit de la Recerca de Catalunya
Publication Artículo científico (JournalArticle)

Acceleration of Born series by change of variables

  • López Menchón, Héctor
  • Rius Casals, Juan Manuel
  • Heldring, Alexander
  • Úbeda Farré, Eduard
In this work, we propose a method to enhance the convergence of the Born series. The Born series is widely used in scattering theory, but its convergence is only guaranteed under certain restrictive conditions which limit the cases where this formulation can be applied. The proposed method, based on modifying the singularities of the resolvent operator by a change of variables, accelerates the convergence of the series or even achieves convergence in otherwise divergent cases. Due to its low computational cost and ease of implementation, the method preserves the advantages of the Born series while it extends its range of application. It can also be applied to a variety of methods relying on the same mathematical principles as the Born series. Numerical examples are provided to show that a significant improvement can be achieved with simple techniques and with a limited knowledge about the operator spectrum., This work was partly funded by the Ministerio de Ciencia e Innovacion (MICINN) under projects TEC2016-78028-C3-1-P, MDM2016-0600, TEC2017-83343-C4-2-R, TEC2017-84817-C2-2-R, PID 2019-107885GB-C31 and Catalan Research Group 2017 SGR 219; and by Secretary of Universities and Research of the Ministry of Business and Knowledge of the Generalitat of Catalonia and European Social Fund under scholarship 2019 FI B00788., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publication Artículo científico (JournalArticle)

Low-rank matrix factorization method for multiscale simulations: a review

  • Li, Mengmeng
  • Ding, Dazhi
  • Heldring, Alexander
  • Hu, Jun
  • Chen, Rushan
  • Vecchi, Giuseppe
In this paper, a review of the low-rank factorization method is presented, with emphasis on their application to multiscale problems. Low-rank matrix factorization methods exploit the rankdeficient nature of coupling impedance matrix blocks between two separated groups. They are widely used, because they are purely algebraic and kernel free. To improve the computation precision and efficiency of low-rank based methods, the improved sampling technologies of adaptive cross approximation (ACA), post compression methods, and the nested low-rank factorizations are introduced. O(N) and O (NlogN) computation complexity of the nested equivalence source approximation can be achieved in low and high frequency regime, which is parallel to the multilevel fast multipole algorithm, N is the number of unknowns. Efficient direct solution and high efficiency preconditioning techniques can be achieved with the low-rank factorization matrices. The trade-off between computation efficiency and time are discussed with respect to the number of levels for low-rank factorizations., This work was supported in part by the Natural Science Foundation of China under Grant 61871222, Grant 62025109, Grant 61890541, Grant 61931021, and Grant 61721001; and in part by the Spanish Government under Project TEC 2017-83343-C4-2-R and Project MDM-2016-0600., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publication Artículo científico (JournalArticle)

Comparison of feature selection techniques for power amplifier behavioral modeling and digital predistortion linearization

  • Barry, Abdoul
  • Li, Wantao
  • Becerra González, Juan Antonio
  • Gilabert Pinal, Pere Lluís
The power amplifier (PA) is the most critical subsystem in terms of linearity and power efficiency. Digital predistortion (DPD) is commonly used to mitigate nonlinearities while the PA operates at levels close to saturation, where the device presents its highest power efficiency. Since the DPD is generally based on Volterra series models, its number of coefficients is high, producing ill-conditioned and over-fitted estimations. Recently, a plethora of techniques have been independently proposed for reducing their dimensionality. This paper is devoted to presenting a fair benchmark of the most relevant order reduction techniques present in the literature categorized by the following: (i) greedy pursuits, including Orthogonal Matching Pursuit (OMP), Doubly Orthogonal Matching Pursuit (DOMP), Subspace Pursuit (SP) and Random Forest (RF); (ii) regularization techniques, including ridge regression and least absolute shrinkage and selection operator (LASSO); (iii) heuristic local search methods, including hill climbing (HC) and dynamic model sizing (DMS); and (iv) global probabilistic optimization algorithms, including simulated annealing (SA), genetic algorithms (GA) and adaptive Lipschitz optimization (adaLIPO). The comparison is carried out with modeling and linearization performance and in terms of runtime. The results show that greedy pursuits, particularly the DOMP, provide the best trade-off between execution time and linearization robustness against dimensionality reduction.</jats:p>, This work was supported in part by the Spanish Government (Ministerio de Ciencia, Inno- vación y Universidades) and Fondo Europeo de Desarrollo Regional (FEDER) under Grants TEC2017- 83343-C4-2-R and PID2020-113832RB-C21 and in part by the Government of Catalonia and the European Social Fund under Grants 2017-SGR-813 and 2021-FI-B-137., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publication Contribución de congreso (ConferenceOutput)

A method for selecting online the coefficients to be updated in a DPD for PA linearization

  • Pham, Thi Quynh Anh
  • Montoro López, Gabriel
  • López Bueno, David
  • Gilabert Pinal, Pere Lluís
This paper presents a technique for selecting onlinethe coefficients to be updated in a digital predistorter (DPD)based on direct learning. The proposed method, which is basedon a combination of matching pursuit (MP) and least squares(LS) techniques (and is therefore named MP-LS method) allowsto improve the power amplifier (PA) linearization performanceof a fixed number of DPD coefficients, due to the fact that ateach DPD iteration the coefficients to be updated are properlychosen. The proposed technique is compared to a conventional LSestimation, and experimental results demonstrate that the MP-LSmethod can provide a performance improvement in relation toa DPD with fixed-preselected coefficients. The method could beespecially useful in DPD systems that have hardware restrictionsin the resources to be used by the update subsystem in thefeedback path. That is the case of DPDs based on FPGA devicesimplementing a QR algorithm in the programmable logic (PL)side, This work was partially supported by the Spanish Govern-ment (MICINN) and FEDER under project TEC2017-83343-C4-2-R, and by the Generalitat de Catalunya under grants 2017SGR 891 and 813., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publication Artículo científico (JournalArticle)

Improving the accuracy of the adaptive cross approximation with a convergence criterion based on random sampling

  • Heldring, Alexander
  • Úbeda Farré, Eduard
  • Rius Casals, Juan Manuel
The accuracy of the Adaptive Cross Approximation (ACA) algorithm, a popular method for the compression of low rank matrix blocks in Method of Moment computations, is sometimes seriously compromised by unpredictable errors in the convergence criterion. This paper proposes an alternative criterion, based on global sampling of the error in the elements of the ACA compressed matrix. The sampling error depends on the size of the sample but also on the population distribution of the error, which makes it difficult to control the error independently of the underlying problem. However, as argued and demonstrated in the paper, the distribution of the error converges to the same unique probability distribution function for all low rank matrices. Complementing the sampling criterion with a simple mechanism to detect this convergence, we arrive at a criterion that controls the error irrespective of the underlying problem. As a practical example the RCS of a moderate size metallic ogive is computed to illustrate the merits of the proposed criterion. The proposed algorithm may also be useful in other methods that approximate low-rank matrices by interpolation of a reduced set of its elements., This work was partly funded by the Ministerio de Ciencia e Innovacion (MICINN) under projects TEC2016-78028-C3-1-P, TEC2017-84817-C2-2-R, TEC2017-83343-C4-2-R, PID 2019-107885GBC31, MDM2016-0600, and Catalan Research Group 2017 SGR 219., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publication Artículo científico (JournalArticle)

Digital predistorters go multidimensional: DPD for concurrent multiband envelope tracking and outphasing power amplifiers

  • Gilabert Pinal, Pere Lluís
  • Montoro López, Gabriel
  • Vegas Garcia, Ángel
  • Ruiz Lavin, Maria de las Nieves
  • García García, José Ángel
© 2019 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works., Over at least the last two decades, digital predistortion (DPD) has become the most common and widespread solution to cope with the power amplifier's (PA's) inherent linearity-versus-efficiency tradeoff. When compared with other linearization techniques, such as Cartesian feedback or feedforward, DPD has proven able to adapt to the always-growing demands of technology: wider bandwidths, stringent spectrum masks, and reconfigurability. The principles of predistortion linearization (in its analog or digital forms) are straightforward, and the linearization subsystem precedes the PA (a nonlinear function in a digital signal processor in the case of DPD or nonlinear device in the case of analog predistortion and counteracts the nonlinear characteristic of the PA. Some excellent overviews on DPD can be found in [1]-[4]. Let us now look at the challenges that DPD linearization has faced and will continue to face in the near future with 5G new radio (5G-NR)., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publication Artículo científico (JournalArticle)

Nested fast adaptive cross approximation algorithm for solving electromagnetic scattering problems

  • Fang, Xiaoxing
  • Heldring, Alexander
  • Rius Casals, Juan Manuel
  • Chen, Xinlei
  • Cao, Qunsheng
In this paper, a Nested Fast Adaptive Cross Approximation (NFACA) algorithm is presented to accelerate the solution of electromagnetic scattering problems. It is based on the Fast Adaptive Cross Approximation (FACA) and the Nested Cross Approximation (NCA). In the NFACA, the FACA is applied instead of the ACA or the Chebyshev nodes, to enhance the efficiency. It is shown to have O(N) storage and computational complexity for electrically small problems. Different from the conventional NCA or the Fast Nested Cross Approximation (FNCA), the pivot-selection does not rely on a preliminary selection with Chebyshev nodes or ACA for each group. The NFACA permits the selection of the optimum pivots independently at each level to make the accuracy controllable. Numerical results of several objects are presented to demonstrate the performance of this method. Compared with the FACA and NCA in solving the electromagnetic scattering problems, the NFACA is more efficient both in computational cost and storage. Compared with the FNCA, the NFACA is both faster and more accurate., This work was supported in part by the National Nature Science Foundation of China under Grant 61871219, Grant 61771238, and Grant 61501227; in part by the Post-Graduate Research and Practice Innovation Program of Jiangsu Province under Grant KYCX18_0289; in part by the China Scholarship Council under Grant 201906830042; in part by the Spanish Ministerio de Ciencia e Innovacion (MICINN) under Project TEC2016-78028-C3-1-P, Project PID 2019-107885GB-C31, Project TEC2017-83343-C4-2-R, and Project MDM2016-0600; and in part by the Catalan Research Group 2017 SGR 219., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publication Contribución de congreso (ConferenceOutput)

Dataset reduction for neural network based digital predistorters under strong nonlinearities

  • López Bueno, David
  • Gilabert Pinal, Pere Lluís
  • Montoro López, Gabriel
The artificial neural networks (ANN) are gaining momentum in the digital predistorters (DPD) thanks to their inherently good approximation capabilities. Under strong or complex power amplifier nonlinearities, the size of the ANN can increase and lead to long training periods which are unaffordable in fast-changing waveform scenarios like those proposed for 5G or 6G. In this work we combine the orthogonal matching pursuit technique together with dataset length reduction methods, to significantly shorten the ANN DPD coefficients update time., This work was supported in part by the Spanish Government (Ministerio de Ciencia, Innovación y Universidades) and the Fondo Europeo de Desarrollo Regional (FEDER) under grants RTI2018-099841-B-I00 and TEC2017-83343-C4-2-R, and in part by the Generalitat de Catalunya under grants 2017 SGR 891 and 2017 SGR 813., Peer Reviewed, Postprint (author's final draft)