• DocumentCode
    983617
  • Title

    Design of a Power-Amplifier Feed-Forward RF Model With Physical Knowledge Considerations

  • Author

    Cunha, Telmo R. ; Pedro, José C. ; Cabral, Pedro M.

  • Author_Institution
    Univ. de Aveiro, Aveiro
  • Volume
    55
  • Issue
    12
  • fYear
    2007
  • Firstpage
    2747
  • Lastpage
    2756
  • Abstract
    In this paper, the design of a general feed-forward topology for modeling RF power amplifiers (PAs) is presented. It consists of producing the best nonlinear nonrecursive approximator of a feedback model which traduces the physical characteristics of general PAs. Following the procedure presented in this paper, the feed-forward structure, which has a polynomial form with memory (Volterra series), can be exactly determined up to the desired order of approximation. Evidently, increasing the order of the feed-forward topology results in a more complex structure, with more laborious model extraction procedures. This paper presents in detail the structures of the general nonrecursive model up to the fifth-order kernel. A systematic extraction procedure, based on standard RF laboratory continuous wave and two-tone tests, is proposed under the assumption that the PA linear output does not vary within the considered fundamental frequency band (flat behavior). Model validation is demonstrated by means of ADS simulation data and from real data measurements.
  • Keywords
    Volterra series; approximation theory; feedforward amplifiers; polynomials; power amplifiers; radiofrequency amplifiers; RF power amplifiers design; Volterra series; feed-forward topology; fifth-order kernel; nonlinear nonrecursive approximation; polynomial form; systematic extraction procedure; Data mining; Feedback; Feedforward systems; Kernel; Laboratories; Polynomials; Power amplifiers; Radio frequency; Radiofrequency amplifiers; Topology; Feed-forward systems; Volterra series; modeling; nonlinear systems; power amplifiers (PAs);
  • fLanguage
    English
  • Journal_Title
    Microwave Theory and Techniques, IEEE Transactions on
  • Publisher
    ieee
  • ISSN
    0018-9480
  • Type

    jour

  • DOI
    10.1109/TMTT.2007.909873
  • Filename
    4385540