Font Size: a A A

Amelioration du rendement energetique des amplicateurs a post-compensation (feedforward) pour signaux a facteur de crete eleve (French text)

Posted on:2003-09-27Degree:Ph.DType:Dissertation
University:Ecole Polytechnique, Montreal (Canada)Candidate:Larose, ColinFull Text:PDF
GTID:1468390011486538Subject:Engineering
Abstract/Summary:
Today's wireless communication networks require base stations that can amplify signals with strongly modulated envelopes. These modulations occur when signals of constant envelope (for example FM) are combined, justifying the need for the so-called multi-carrier power amplifier (MCPA). Whether combined or not, signals may also have inherent envelope modulation, as in the CDMA (Code Division Multiple Access) digital cellular system.; The present work begins with the development of an adequate numerical method of optimizing the feedforward amplifier. The procedure is based on a realistic model in which the vector modulators precede the main and error amplifiers (which influences how the adaptation methods perform) and the relationships governing the various losses, delays, and tolerances in the circuit are taken into account. By way of seven design parameters, the procedure optimizes power efficiency in accordance with the customer's gain and linearity requirements, and not the other way around. The complex baseband analysis makes for better computational efficiency and allows taking into account the AM-PM conversion, which will prove to be central to the phenomena under study. Although narrowband, this analysis proved adequate in predicting the spectral regrowth of a CDMA signal exiting a typical main amplifier. The statistical approach further accelerates the computations and merely requires of the input signal a statistical description derived from the complementary cumulative distribution function (CCDF), whose use is widespread. Finally, the non-analytical relationships are effectively solved by means of a particularly intuitive graphical representation.; The feedforward optimization procedure cannot directly optimize the type of amplifier and the adaptation method employed in each loop. Hence, the procedure requires from the outset four normalized functions to convey the gain compression and dc power characteristics of the two amplifiers. These characteristics depend on the type of transistor and the class of amplification as determined by the biasing of the transistor.; The work concludes with efforts to optimize the types of amplifiers and the adaptation method to be used in the feedforward amplifier. (Abstract shortened by UMI.)...
Keywords/Search Tags:Feedforward, Amplifier
Related items