Please use this identifier to cite or link to this item:
A Study of Universal Formulas for Nonlinear Channel Identification in Orthogonal Frequency-Division Multiplexing Systems
|Contributors: ||NTOU:Department of Electrical Engineering|
OFDM;nonlinear channel;channel estimation;Volterra kernel;system identifica-tion;higher-order auto-moment spectrum
|Issue Date: ||2013-10-07T02:32:43Z
In this project, we propose to derive universal formulas for identifying arbitrary-order nonlinear channels in orthogonal frequency-division multiplexing (OFDM) systems. The OFDM is a well-developed modulation scheme for wideband communication applications. It has been adopted by many wireline and wireless communication standards to achieve high-speed data transmissions. It is well known that the OFDM signal could be very sensitive to nonlinearities of power amplifiers in a communication link. Therefore, estimation of the nonlinear communication channel is essential to the design of an OFDM system. One com-monly used model for characterizing nonlinear channels is the Volterra series. A nonlinear bandpass channel is commonly modeled by a baseband equivalent Volterra series which re-lates the complex envelopes of the channel input and output. Modeling the nonlinear channel requires the identification of the time-domain or the frequency-domain Volterra kernels of the channel. The number of the required kernel coefficients is often very large even for a rela-tively low-order Volterra model. This makes the identification of the Volterra kernels a very challenging task, especially when the order of the nonlinearities is large. Various methods for identifying the Volterra kernels of nonlinear channels have been proposed in the literature. These methods, although are suitable for certain occasions with relatively low order nonli-nearities, might not be easily extendable to be applied to cases with higher-order nonlineari-ties in a straightforward manner. We have recently derived a novel method to estimate the frequency-domain baseband equivalent Volterra kernels of cubically nonlinear bandpass channels in OFDM systems. By exploring the higher-order auto-moment spectral properties of the OFDM signal, this method gives the closed-form solution for the frequency-domain baseband equivalent Volterra kernels of the nonlinear channel up to the third order. It can be used to identify cubically nonlinear channels in OFDM systems which employ QAM or PSK modulations in their subchannels. The obtained kernel estimates by this method are optimal in the minimum mean square error (MMSE) sense. A limitation of this method is that it is appli-cable only up to third-order nonlinear OFDM systems. The goal of this project is to perform an advanced and systematic study on generalizing the method we have derived for cubically nonlinear OFDM systems, so that it can be applicable to nonlinear OFDM systems of any or-der. Specifically, we will derive universal formulas of closed-form expressions for the fre-quency-domain Volterra kernels of nonlinear OFDM systems with an arbitrary order. The de-rived universal formulas will theoretically give the optimal MMSE solution of the Volterra kernels. Moreover, techniques for guaranteeing the estimated Volterra kernels to attain their optimal the MMSE solution will also be investigated. The correctness of the derived universal formulas and the effectiveness of the proposed method will be justified by computer simula-tion and real experiment. The completion of this project can consolidate the theory and appli-cation of the Volterra series for nonlinear channel identification in OFDM systems.
|Appears in Collections:||[電機工程學系] 研究計畫|
Files in This Item:
There are no files associated with this item.
All items in NTOUR are protected by copyright, with all rights reserved.