Pasar al contenido principal

DISPOSITIVOS ASISTIDOS POR TECNICAS DE DEEP Y MACHINE LEARNING PARA TRANSCEPTORES DE RADIOFRECUENCIA ALTAMENTE EFICIENTES

PID2020-113832RB-C21

•
Nombre agencia financiadora Agencia Estatal de Investigación
Acrónimo agencia financiadora AEI
Programa Programa Estatal de I+D+i Orientado a los Retos de la Sociedad
Convocatoria Proyectos de I+D+i (Generación de Conocimiento y Retos Investigación)
Año convocatoria 2020
Plan Estatal 2017-2020 Plan Estatal de Investigación Científica y Técnica y de Innovación
Centro beneficiario UNIVERSITAT POLITECNICA DE CATALUNYA

Resultados relacionados

Resultados totales (Incluyendo duplicados): 18
Encontrada(s) 1 página(s)

UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Frequency-Dependent DPD linearization for load-mismatched mobile terminal PAs operating under dynamic resource block allocation

  • Li, Wantao|||0000-0002-2634-6742
  • Montoro López, Gabriel|||0000-0002-1328-4175
  • Guo, Yan
  • Gilabert Pinal, Pere Lluís|||0000-0001-6183-6977
This paper presents a novel linearization method to compensate for load-mismatched power amplifiers in 5G new radio mobile terminals. The proposed digital predistortion (DPD) approach is based on a frequency-dependent memoryless polynomial behavioral model that allows meeting the linearity specifications under dynamic resource block (RB) allocation when considering narrowband signals. The advantage of the proposed frequency-dependent method is experimentally validated by comparing it with a classical DPD approach where the coefficients are trained only for a specific RB location., This project was funded by Huawei Technologies from July 2021 to August 2022; and supported in part by the Ministerio de Ciencia e Innovación and FEDER under Grant PID2020-113832RB-C21 and in part by the Government of Catalonia and the European Social Fund under Grant 2021-FI-B-137., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Improvements in randomized matrix compression methods for numerical analysis of antennas and scatterers

  • López Menchón, Héctor|||0000-0001-5776-1423
  • Heldring, Alexander|||0000-0003-2011-2096
  • Úbeda Farré, Eduard|||0000-0001-6759-0445
  • Rius Casals, Juan Manuel|||0000-0003-0606-5422
We propose a randomized Pseudo Skeleton (or CUR) Approximation method to compress the H-matrices of linear systems that arise in the discretization of integral equations in electromagnetic scattering. This method is highly parellelizable and well suited for fine-granularity architectures as GPUs. As the Adaptive Approximation Method (ACA), the method is purely algebraic. The method is tested with standard cases to assess its quality and efficiency, This work was partly funded by the Ministerio de Ciencia e Innovacion (MICINN) under projects PID2019-107885GBC31 / AEI / 10.13039/501100011033, PID2020-113832RBC21 / AEI / 10.13039/501100011033 and PID-2020-118410RB-C21 / AEI / 10.13039/501100011033, PDC2022-133091-I00 / AEI / 10.13039/501100011033, and Catalan Research Group 2017 SGR 219 and grant 2021 FI B2 00096., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Simple randomized strategies for low-rank matrix compression

  • López Menchón, Héctor|||0000-0001-5776-1423
  • Molina Burgués, Robert|||0009-0005-1314-3742
  • Heldring, Alexander|||0000-0003-2011-2096
  • Úbeda Farré, Eduard|||0000-0001-6759-0445
  • Rius Casals, Juan Manuel|||0000-0003-0606-5422
We propose a randomized Pseudo Skeleton (or CUR) approximation method to compress the H-matrices of linear systems that arise in the discretization of integral equations in electromagnetic scattering. This method is essentially the Matrix Decomposition Algorithm (MDA) of Michielssen and Boag (1994) with equivalent basis and testing functions equal to a random subset of the original ones. It is highly parellelizable and well suited for fine-granularity architectures as GPUs. As the Adaptive Cross Approximation (ACA), CUR is purely algebraic. The method is tested with standard cases to assess its quality and efficiency., This work was partly funded by the Ministerio de Ciencia e Innovacion (MICINN) under projects PID2019-107885GBC31 / AEI / 10.13039/501100011033, PID2022-136869NBC31 / AEI / 10.13039/501100011033, PID2020-113832RBC21 / AEI / 10.13039/501100011033 and PID-2020-118410RB-C21 / AEI / 10.13039/501100011033, PDC2022-133091-I00 / AEI / 10.13039/501100011033, and Catalan Research Group 2017 SGR 219 and grant 2021 FI B2 00096., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

ANN versus cascaded behavioral models for DPD linearization of wideband dual-input PAs

  • Criado Simón, Raúl
  • Li, Wantao|||0000-0002-2634-6742
  • Montoro López, Gabriel|||0000-0002-1328-4175
  • Thompson, William
  • Chuang, Kevin
  • Gilabert Pinal, Pere Lluís|||0000-0001-6183-6977
© 2025 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., This paper offers a thorough comparison of two families of digital predistortion (DPD) behavioral models used for the linearization of high-efficiency dual-input wideband power amplifiers. Specifically, N-stage cascaded (CC) models are compared with artificial neural networks (ANNs) models in terms of linearization performance and computational complexity. The DPD models are designed balancing the competing demands of linearity, output power, and power efficiency of a pseudoDoherty load-modulated balanced amplifier (PD-LMBA), while also considering the complexity of the model. Experimental results using a 5G new radio test signal with 200 MHz instantaneous bandwidth demonstrate the potential benefits of CC and ANN behavioral models to linearize a high-efficiency PD-LMBA at an RF frequency of 2 GHz, delivering a mean output power level of 40 dBm at around 50% power efficiency., The work in this paper was supported in part by the project PID2020113832RB-C21 funded by MCIN/AEI/10.13039/50110001103 and in part by the Gov. of Catalonia and the Europ. Social Fund under Grant 2021-FI-B-137.
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Solving electromagnetic scattering in dielectric bodies by Monte Carlo sampling

  • López Menchón, Héctor|||0000-0001-5776-1423
  • Rius Casals, Juan Manuel|||0000-0003-0606-5422
  • Úbeda Farré, Eduard|||0000-0001-6759-0445
  • Heldring, Alexander|||0000-0003-2011-2096
In this work, we propose Monte Carlo sampling as a technique to avoid communication overhead in the numerical solution of complex electromagnetic scattering problems. The method is based on statistically sampling a formal representation of the solution of the scattering problem, based on a modified Born series. The sampling process can be independently carried out in different processors with no need for communication between them. This lack of communication makes the algorithm suitable for massively parallel computational environments., This work was partly funded by the Ministerio de Ciencia e Innovacion (MICINN) under projects PID2019-107885GB-C31 / AEI / 10.13039/501100011033, PID2020-113832RB- C21 / AEI / 10.13039/501100011033 and TEC2017- 84817-C2-2-R / AEI / 10.13039/501100011033 and the Unidad de Excelencia Maria de Maeztu MDM-2016- 0600/AEI/10.13039/501100011033, which is financed by the Agencia Estatal de Investigaci ´on, Spain, and Catalan Research Group 2017 SGR 219, and Ag`encia de Gesti ´o d’Ajuts Uni- versitaris i de Recerca with grant 2021 FI B2 00096., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Incremental DPD linearization for mobile terminals with non-flat frequency response in dynamic bandwidth Re-allocation scenarios

  • Li, Wantao|||0000-0002-2634-6742
  • Guo, Yan
  • Montoro López, Gabriel|||0000-0002-1328-4175
  • Gilabert Pinal, Pere Lluís|||0000-0001-6183-6977
This paper presents a low-complexity open-loop digital linearization system for handset applications with non-flat frequency response when considering a 5G new radio dynamic bandwidth re-allocation scenario. The proposed digital predistorter (DPD) is scalable with the signal bandwidth by simply activating or deactivating basis kernels, without re-training nor re-assigning the model coefficients. To support this feature, an incremental bandwidth (IBW) generalized memory polynomial behavioral model is proposed, and the coefficients adaptation procedure is described. In order to find the most relevant basis supporting the IBW feature, a constrained version of the doubly orthogonal matching pursuit (DOMP) algorithm is proposed. Experimental results considering a handset power amplifier (PA) SoC under a mismatched load condition (i.e., non-flat frequency response) will validate the proposed approach., The work in this paper was funded by Huawei Technologies from January 2022 to December 2022; and supported in part by the project PID2020-113832RB-C21 funded by MCIN/AEI/10.13039/50110001103 and in part by the Government of Catalonia and the European Social Fund under Grant 2021-FI-B-137., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Design and implementation of a GPU-Based digital predistortion linearizer for RF power amplifiers

  • Li, Wantao|||0000-0002-2634-6742
  • Montoro López, Gabriel|||0000-0002-1328-4175
  • Gilabert Pinal, Pere Lluís|||0000-0001-6183-6977
This paper presents an automated process for generating computationally efficient radiofrequency (RF) power amplifier (PA) behavioural models that can be used for digital-predistortion (DPD) linearization. The implementation of the DPD behavioral model is based on the personal computer (PC) architecture including recent graphical process units (GPUs) with compute unified device architecture (CUDA) cores. By utilizing the parallel computing capability, the proposed implementation can achieve high baseband throughput required by nowadays 5G new radio (NR) communications and the inherent DPD bandwidth expansion. The details of the DPD implementation will be given and the advantages and disadvantages of the GPU DPD will be discussed. Experimental results will validate the DPD implementation in a hardware-in-the-loop (HITL) environment with a system-on-chip (SoC) PA operated with a 100 MHz 5G-NR signal., The work in this paper was funded by Huawei Technologies from January 2022 to December 2022; and supported in part by the project PID2020- 113832RB-C21 funded by MCIN/AEI/10.13039/50110001103 and in part by the Government of Catalonia and the European Social Fund under Grant 2021- FI-B-137, Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Digital shaping and linearization of a dual-input load-modulated balanced amplifier

  • Li, Wantao|||0000-0002-2634-6742
  • Montoro López, Gabriel|||0000-0002-1328-4175
  • Thompson, William
  • Chuang, Kevin
  • Gilabert Pinal, Pere Lluís|||0000-0001-6183-6977
This paper presents shaping and linearization techniques to trade-off linearity and power efficiency in a dual-input pseudo-Doherty load-modulated balanced amplifier (PD-LMBA). The proposed shaping function is defined to operate in the frequency domain and it is oriented at exploiting the dual-input architecture of the PD-LMBA to either enhance linearity or power efficiency. In addition, since the linearity requirements are not met by simply shaping the input signals, digital predistortion (DPD) linearization is applied on top of it. Experimental validation of the proposed shaping function and DPD linearization is provided for a PD-LMBA excited with 200 MHz bandwidth 5G-like modulated signals with 8 dB of PAPR. For an output power around 40 dBm, the PD-LMBA presents a power efficiency around 50%, while meeting the linearity specs of -45 dB of adjacent channel power ratio (ACPR) and error vector magnitude (EVM) below 3%., The work in this paper was supported in part by the project PID2020- 113832RB-C21 funded by MCIN/AEI/10.13039/50110001103 and in part by the Government of Catalonia and the European Social Fund under Grant 2021- FI-B-137., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Fast modeling of lateral modes in BAW resonators with arbitrary in-plane geometry

  • Udaondo Guerrero, Carlos|||0000-0002-4251-0530
  • Collado Gómez, Juan Carlos|||0000-0002-8869-2739
  • Mateu Llevadot, Jordi|||0000-0003-1735-9434
The Transmission Line Matrix (TLM) method is used to model and predict lateral spurious resonances in a BAW resonator. The dispersive characteristics of the main propagating mode, which are defined by the stack of materials, are considered with dispersive transmission lines along the in-plane lateral directions of the resonator. The particle displacement profiles obtained with the TLM method are coupled to the electric domain. From these, the input impedance of the resonator is obtained. This allows for fast simulations of the lateral resonances using nodal admittance matrices. © 2022 IEEE., This work was supported in part by the Catalan Government under Grant 2017 SGR 813, in part by the Spanish Government under Grant PID2020- 118410RB-C21 and Grant PID2020-113832RB-C21, and was also supported by the Secretary of Universities and Research of the Generalitat de Catalunya and the European Social Fund through grant 2020 FI_B-00967, Peer Reviewed
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Improvements in randomized matrix compression methods for numerical analysis of antennas and scatterers

  • López Menchón, Héctor
  • Heldring, Alexander
  • Úbeda Farré, Eduard
  • Rius Casals, Juan Manuel
We propose a randomized Pseudo Skeleton (or CUR) Approximation method to compress the H-matrices of linear systems that arise in the discretization of integral equations in electromagnetic scattering. This method is highly parellelizable and well suited for fine-granularity architectures as GPUs. As the Adaptive Approximation Method (ACA), the method is purely algebraic. The method is tested with standard cases to assess its quality and efficiency, This work was partly funded by the Ministerio de Ciencia e Innovacion (MICINN) under projects PID2019-107885GBC31 / AEI / 10.13039/501100011033, PID2020-113832RBC21 / AEI / 10.13039/501100011033 and PID-2020-118410RB-C21 / AEI / 10.13039/501100011033, PDC2022-133091-I00 / AEI / 10.13039/501100011033, and Catalan Research Group 2017 SGR 219 and grant 2021 FI B2 00096., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Simple randomized strategies for low-rank matrix compression

  • López Menchón, Héctor
  • Molina Burgués, Robert
  • Heldring, Alexander
  • Úbeda Farré, Eduard
  • Rius Casals, Juan Manuel
We propose a randomized Pseudo Skeleton (or CUR) approximation method to compress the H-matrices of linear systems that arise in the discretization of integral equations in electromagnetic scattering. This method is essentially the Matrix Decomposition Algorithm (MDA) of Michielssen and Boag (1994) with equivalent basis and testing functions equal to a random subset of the original ones. It is highly parellelizable and well suited for fine-granularity architectures as GPUs. As the Adaptive Cross Approximation (ACA), CUR is purely algebraic. The method is tested with standard cases to assess its quality and efficiency., This work was partly funded by the Ministerio de Ciencia e Innovacion (MICINN) under projects PID2019-107885GBC31 / AEI / 10.13039/501100011033, PID2022-136869NBC31 / AEI / 10.13039/501100011033, PID2020-113832RBC21 / AEI / 10.13039/501100011033 and PID-2020-118410RB-C21 / AEI / 10.13039/501100011033, PDC2022-133091-I00 / AEI / 10.13039/501100011033, and Catalan Research Group 2017 SGR 219 and grant 2021 FI B2 00096., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Digital shaping and linearization of a dual-input load-modulated balanced amplifier

  • Li, Wantao
  • Montoro López, Gabriel
  • Thompson, William
  • Chuang, Kevin
  • Gilabert Pinal, Pere Lluís
This paper presents shaping and linearization techniques to trade-off linearity and power efficiency in a dual-input pseudo-Doherty load-modulated balanced amplifier (PD-LMBA). The proposed shaping function is defined to operate in the frequency domain and it is oriented at exploiting the dual-input architecture of the PD-LMBA to either enhance linearity or power efficiency. In addition, since the linearity requirements are not met by simply shaping the input signals, digital predistortion (DPD) linearization is applied on top of it. Experimental validation of the proposed shaping function and DPD linearization is provided for a PD-LMBA excited with 200 MHz bandwidth 5G-like modulated signals with 8 dB of PAPR. For an output power around 40 dBm, the PD-LMBA presents a power efficiency around 50%, while meeting the linearity specs of -45 dB of adjacent channel power ratio (ACPR) and error vector magnitude (EVM) below 3%., The work in this paper was supported in part by the project PID2020- 113832RB-C21 funded by MCIN/AEI/10.13039/50110001103 and in part by the Government of Catalonia and the European Social Fund under Grant 2021- FI-B-137., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Incremental DPD linearization for mobile terminals with non-flat frequency response in dynamic bandwidth Re-allocation scenarios

  • Li, Wantao
  • Guo, Yan
  • Montoro López, Gabriel
  • Gilabert Pinal, Pere Lluís
This paper presents a low-complexity open-loop digital linearization system for handset applications with non-flat frequency response when considering a 5G new radio dynamic bandwidth re-allocation scenario. The proposed digital predistorter (DPD) is scalable with the signal bandwidth by simply activating or deactivating basis kernels, without re-training nor re-assigning the model coefficients. To support this feature, an incremental bandwidth (IBW) generalized memory polynomial behavioral model is proposed, and the coefficients adaptation procedure is described. In order to find the most relevant basis supporting the IBW feature, a constrained version of the doubly orthogonal matching pursuit (DOMP) algorithm is proposed. Experimental results considering a handset power amplifier (PA) SoC under a mismatched load condition (i.e., non-flat frequency response) will validate the proposed approach., The work in this paper was funded by Huawei Technologies from January 2022 to December 2022; and supported in part by the project PID2020-113832RB-C21 funded by MCIN/AEI/10.13039/50110001103 and in part by the Government of Catalonia and the European Social Fund under Grant 2021-FI-B-137., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

ANN versus cascaded behavioral models for DPD linearization of wideband dual-input PAs

  • Criado Simón, Raúl
  • Li, Wantao
  • Montoro López, Gabriel
  • Thompson, William
  • Chuang, Kevin
  • Gilabert Pinal, Pere Lluís
© 2025 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., This paper offers a thorough comparison of two families of digital predistortion (DPD) behavioral models used for the linearization of high-efficiency dual-input wideband power amplifiers. Specifically, N-stage cascaded (CC) models are compared with artificial neural networks (ANNs) models in terms of linearization performance and computational complexity. The DPD models are designed balancing the competing demands of linearity, output power, and power efficiency of a pseudoDoherty load-modulated balanced amplifier (PD-LMBA), while also considering the complexity of the model. Experimental results using a 5G new radio test signal with 200 MHz instantaneous bandwidth demonstrate the potential benefits of CC and ANN behavioral models to linearize a high-efficiency PD-LMBA at an RF frequency of 2 GHz, delivering a mean output power level of 40 dBm at around 50% power efficiency., The work in this paper was supported in part by the project PID2020113832RB-C21 funded by MCIN/AEI/10.13039/50110001103 and in part by the Gov. of Catalonia and the Europ. Social Fund under Grant 2021-FI-B-137., Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Fast modeling of lateral modes in BAW resonators with arbitrary in-plane geometry

  • Udaondo Guerrero, Carlos
  • Collado Gómez, Juan Carlos
  • Mateu Llevadot, Jordi
The Transmission Line Matrix (TLM) method is used to model and predict lateral spurious resonances in a BAW resonator. The dispersive characteristics of the main propagating mode, which are defined by the stack of materials, are considered with dispersive transmission lines along the in-plane lateral directions of the resonator. The particle displacement profiles obtained with the TLM method are coupled to the electric domain. From these, the input impedance of the resonator is obtained. This allows for fast simulations of the lateral resonances using nodal admittance matrices. © 2022 IEEE., This work was supported in part by the Catalan Government under Grant 2017 SGR 813, in part by the Spanish Government under Grant PID2020- 118410RB-C21 and Grant PID2020-113832RB-C21, and was also supported by the Secretary of Universities and Research of the Generalitat de Catalunya and the European Social Fund through grant 2020 FI_B-00967, Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Solving electromagnetic scattering in dielectric bodies by Monte Carlo sampling

  • López Menchón, Héctor
  • Rius Casals, Juan Manuel
  • Úbeda Farré, Eduard
  • Heldring, Alexander
In this work, we propose Monte Carlo sampling as a technique to avoid communication overhead in the numerical solution of complex electromagnetic scattering problems. The method is based on statistically sampling a formal representation of the solution of the scattering problem, based on a modified Born series. The sampling process can be independently carried out in different processors with no need for communication between them. This lack of communication makes the algorithm suitable for massively parallel computational environments., This work was partly funded by the Ministerio de Ciencia e Innovacion (MICINN) under projects PID2019-107885GB-C31 / AEI / 10.13039/501100011033, PID2020-113832RB- C21 / AEI / 10.13039/501100011033 and TEC2017- 84817-C2-2-R / AEI / 10.13039/501100011033 and the Unidad de Excelencia Maria de Maeztu MDM-2016- 0600/AEI/10.13039/501100011033, which is financed by the Agencia Estatal de Investigaci ´on, Spain, and Catalan Research Group 2017 SGR 219, and Ag`encia de Gesti ´o d’Ajuts Uni- versitaris i de Recerca with grant 2021 FI B2 00096., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Frequency-Dependent DPD linearization for load-mismatched mobile terminal PAs operating under dynamic resource block allocation

  • Li, Wantao
  • Montoro López, Gabriel
  • Guo, Yan
  • Gilabert Pinal, Pere Lluís
This paper presents a novel linearization method to compensate for load-mismatched power amplifiers in 5G new radio mobile terminals. The proposed digital predistortion (DPD) approach is based on a frequency-dependent memoryless polynomial behavioral model that allows meeting the linearity specifications under dynamic resource block (RB) allocation when considering narrowband signals. The advantage of the proposed frequency-dependent method is experimentally validated by comparing it with a classical DPD approach where the coefficients are trained only for a specific RB location., This project was funded by Huawei Technologies from July 2021 to August 2022; and supported in part by the Ministerio de Ciencia e Innovación and FEDER under Grant PID2020-113832RB-C21 and in part by the Government of Catalonia and the European Social Fund under Grant 2021-FI-B-137., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Design and implementation of a GPU-Based digital predistortion linearizer for RF power amplifiers

  • Li, Wantao
  • Montoro López, Gabriel
  • Gilabert Pinal, Pere Lluís
This paper presents an automated process for generating computationally efficient radiofrequency (RF) power amplifier (PA) behavioural models that can be used for digital-predistortion (DPD) linearization. The implementation of the DPD behavioral model is based on the personal computer (PC) architecture including recent graphical process units (GPUs) with compute unified device architecture (CUDA) cores. By utilizing the parallel computing capability, the proposed implementation can achieve high baseband throughput required by nowadays 5G new radio (NR) communications and the inherent DPD bandwidth expansion. The details of the DPD implementation will be given and the advantages and disadvantages of the GPU DPD will be discussed. Experimental results will validate the DPD implementation in a hardware-in-the-loop (HITL) environment with a system-on-chip (SoC) PA operated with a 100 MHz 5G-NR signal., The work in this paper was funded by Huawei Technologies from January 2022 to December 2022; and supported in part by the project PID2020- 113832RB-C21 funded by MCIN/AEI/10.13039/50110001103 and in part by the Government of Catalonia and the European Social Fund under Grant 2021- FI-B-137, Peer Reviewed, Postprint (published version)