Nonlinearity compensation in digital radio systems

Digital radio links with bandlimited pulses exhibit a severe performance degradation when the transmitter high power amplifier operates near saturation. To cope with the increase of nonlinear intersymbol interference due to the amplifier nonlinearities, a discrete-time Volterra system can be used to process the transmitted data. We present an efficient technique for implementing adaptive data predistorters with memory based on a discretetime Volterra system composed of digital linear filters and memoryless nonlinear devices working at the symbol rate. Thirdand fifth-order structures are proposed and a system performance evaluation is presented for several realistic situations. I . INTRODUCTION IN order to allow the maximum exploitation of the RF spectrum, the current trend in the design of digital radio systems is to increase the spectrum efficiency by using high capacity modulation formats such as multilevel quadrature amplitude modulation (&AM) with bandlimited pulses to reduce the adjacent channel interference (ACI). On the other hand, the need €or good power efficiency forces the drawing of the transmitter high power amplifier (HPA) to near saturation, where the nonlinear distortion, acting on a bandlimited pulse stream, gives rise to some unwanted effects, such as: i) the increase of the intersymbol interference (ISI) (the detector receives a warped constellation of clusters), and ii) the increase of the ACI due to a widening of the transmitted signal spectrum. The severity of both these effects increases as the size of the alphabet increases. When the HPA nonlinearity becomes a significant source of impairment a number of questions arise. For instance: what special circuitry should be used to compensate for the nonlinearity? What is the best design for the filter shaping? What are the best constellations? In recent years many authors have given some partial solutions to these problems. The proposed methods may be divided into two main classes, those operating on the transmitter side (TX), and those operating on the receiver side (RX). The optimal solution among the RX-techniques is the maximum likelihood sequence estimation (MLSE) technique proposed by Mesiya, Mc Lane and Campbell [1] and also by Van Etten Paper approved by Peter J. Mc Lane, the Editor for Modulation Theory and Nonlinear Channels of IEEE Communications Society. Manuscript received March 2, 1990; revised April 18, 1991; January 28, 1992. This work was supported in part by MURST, the Italian Ministry of the University and Scientific Research. S. Pupolin and G. Lazzarin are with the Dipartimento di Elettronica e Informatica, Universit i di Padova, via Gradenigo 6/A, 35131 Padua, Italy, A. Sarti is with the Dipartimento di Elettronica e Informazione, Politecnico di Milano, Piazza L D a Vinci 32, 20133 Milan, Italy. IEEE Log Number 9400996. 0090-6778/94$04.00