This proposal describes a systematic investigation into the fundamental mechanisms that contribute to radio frequency (RF) power-amplifier spectral regrowth, a critical issue for the wireless communications industry. These mechanisms include transient (i.e. memory) effects associated with trapping and phase-distortion due to the intrinsic transistor device structure, input/output match, and bias network impedances. Previously reported work in this area is incomplete, fragmented, and contradictory. The proposed research is divided into several easily identifiable tasks:

Task I Detailed modeling and analysis ofFET phase distortion Task II Investigation into transient memory effects on spectral regrowth Task III Wide-band behavioral signal modeling development Task I is intended to determine the cause of phase distortion in the FET, an issue that is still hotly debated. Task II, which comprises the majority of our total effort, will leverage the recent work of Leoni, et. al. at Lehigh University which addressed transient effects in a GaAs FET power-amplifier under pulse modulation excitation[I]. This work will be extended to include, for the first time, the effects of transients on the output spectrum of a complex signal such as Code Division Multiple Access (CDMA). Spectral regrowth effects associated with input/output match and bias circuit impedance )which have not been systematically studied to date) will also be investigated. The goals of Task H are to develop a fundamental understanding of the effects of transients on spectral regrowth generation and to develop design techniques to mitigate these effects. Task IH leverages recent behavioral modeling work at Lafayette College [2]. The goal of this task is to develop a behavioral model that can be applied to new waveforms such as wide-band CDMA.

The knowledge gained from these investigations will be applied to develop better large-signal transistor models and behavioral circuit models that can handle complex digitally modulated signals such as CDMA. These models will be validated by novel and extensive measurements using devices provided by our industrial partner, Anadigics, Inc., a leading worldwide supplier of RF power-amplifiers for the wireless communications industry. The results of this work will lead to the development of improved RF power-amplifier modeling and design procedures for digital communication applications. ***

Project Start
Project End
Budget Start
1999-06-01
Budget End
2002-05-31
Support Year
Fiscal Year
1999
Total Cost
$175,290
Indirect Cost
Name
Lafayette College
Department
Type
DUNS #
City
Easton
State
PA
Country
United States
Zip Code
18042