• DocumentCode
    568833
  • Title

    A novel predistorter for highly nonlinear broadband power amplifier

  • Author

    Bruno, Marcelo ; Gregorio, Fernando ; Cousseau, Juan ; Ghadam, Ali Shahed Hagh ; Valkama, Mikko

  • fYear
    2012
  • fDate
    9-10 Aug. 2012
  • Firstpage
    84
  • Lastpage
    89
  • Abstract
    In this work we propose a partitioned predistorter (PPD) to compensate the nonlinear impairments produced by a Class AB LDMOS high power amplifier (PA). This kind of PA exhibits distinct input-output characteristics depending on the power regions, making difficult to compensate the distortion with a conventional predistorter (PD). A novel memory polynomial based PPD is implemented in order to linearize both the gain expansion and compression at low and high power regimes. Each stage of the DPD works in a different input amplitude region, allowing the reduction of polynomial order needed to compensate the overall nonlinear characteristic. The performance of the proposed PPD is validated by simulations. Normalized mean squared error (NMSE) and adjacent channel power ratio (ACPR) are evaluated showing significant improvement compared with conventional PD systems.
  • Keywords
    mean square error methods; power amplifiers; wideband amplifiers; ACPR; NMSE; adjacent channel power ratio; amplitude region; class-AB LDMOS high-power amplifier; gain compression; gain expansion; highly-nonlinear broadband power amplifier; input-output characteristics; memory polynomial-based PPD; nonlinear impairment compensation; normalized mean squared error; partitioned predistorter; polynomial order reduction; Educational institutions; Estimation; Nonlinear distortion; OFDM; Polynomials; Predistortion; Radio frequency; LDMOS; WiMAX; efficiency; linearization; memory effects; memory polynomial model; partitioned predistorter;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Micro-Nanoelectronics, Technology and Applications (EAMTA), 2012 Argentine School of
  • Conference_Location
    Cordoba
  • Print_ISBN
    978-1-4673-2696-4
  • Type

    conf

  • Filename
    6297323