Digital Signal Processing 20 (2010) 191–200 Contents lists available at ScienceDirect Digital Signal Processing www.elsevier.com/locate/dsp Imperfect cross-correlation and amplitude balance effects on conventional multiuser decoder with turbo encoding Simona V. Halunga, Nicolae Vizireanu, Octavian Fratu ∗ University “Politehnica” of Bucharest, Electronics, Telecommunications & Information Technology Faculty, Telecommunications Dept., Iuliu Maniu, 1-3, 061071 Bucharest, Romania article info abstract Article history: Available online 22 May 2009 Keywords: Imperfect detection conditions Multiuser detection Performance evaluation turbo coding The present work aims to illustrate the improvement that turbo encoding/decoding algorithms can bring to a multiuser detection system, evaluated in terms of bit error rate (BER). Extended simulations have been performed for conventional multiuser detection algorithms and are still under study for other algorithms, like the optimal, the MMSE and the classical MMSE ones. Simulations have been performed in both optimal conditions and imperfect ones, like non-equal power of users and non-zero correlations between them and for different number of users. Several interesting conclusions have been highlighted with respect to parameters influence on overall system performances.  2009 Elsevier Inc. All rights reserved. 1. Introduction The theory multiuser detection technique has been developed during the 90s [1], but the interest on the domain in- creased with the development of modern mobile communications systems and the application of this technique, especially at base station level, where all the signals have to be demodulated [2,3]. The domain is still under research, since it has not been found yet an optimal solution that maximizes the performances, minimizes the cost and takes into account the chan- nel impairments and system imperfections. The interest on the multiuser detection increased in the mobile communications systems beyond 3G, due the extensive use of MIMO technique. In this case both the base stations and the mobile terminals can have better performances using multiuser detection for the joint received signals [4,5]. Several research studies have been made in order to determine less complex receiver algorithms and to determine their performances under different propagation conditions [4,6]. Iterative decoding algorithms based on turbo decoding principles have also been studied [7,8], as well as evaluation of performances in different fading channel conditions [9] and in combination with space time coding [10]. Turbo decoding, having performances closed to the theoretical Shannon limit [11,12], has several complementary perfor- mances in comparison to the multiuser detection. The turbo equalization has been shown robust and efficient to combat the selective fading, by utilizing the soft decisions on signals for iterative equalization and channel estimation [13,14]. Some recent papers [13,15] highlight that the use of frequency-domain processing decreases the complexity of detection algo- rithms. Moreover, several versions are proposed according to the used type of multiple access (FDMA [13], TDMA [15,16], CDMA [17] or OFDM [18]) and all of them represent enhancements for the existing mobile communications systems. The interleaving procedure [19] used in turbo coding could be optimized and enhanced [20]. * Corresponding author. E-mail address: octavianfratu@gmail.com (O. Fratu). 1051-2004/$ – see front matter  2009 Elsevier Inc. All rights reserved. doi:10.1016/j.dsp.2009.05.007