Analytical performance and improvement for MIMO MC-CDMA systems

Abstract

In this thesis, we evaluate and improve the detection techniques for MC-CDMA systems. For the basic MC-CDMA system, we provide the BER expression for orthogonality restoring combining (ORC) and minimum mean square error combining (MMSEC). ORC can perfectly get rid of the effect of the frequency selective channel which leads to the elimination of the multiple access interference (MAI), generated in the process of user multiplexing process. Thus, for usually high loaded condition, ORC scheme has an advantage to apply. For deriving the performance of ORC, we use the statistics of the combining output signal. Using the moment generating function (MGF) of the combined fading random variable, we present an exact and a simple approximation bit error rate (BER) expression of ORC over Nakagami-$\It{m}$ fading channel. Also we investigate the performance of MMSEC over Rayleigh fading channel. Since the distribution of combined signal has not been derived in the literature, we present a simple lower bound and an approximated expression on the BER. According to the request of high data rate transmission, MC-CDMA systems combine multi-antenna techniques, called multi-input multi-output (MIMO), to obtain higher spectral efficiency. In MC-CDMA systems, the multiplexed signal cannot be directly separated at the receiver since there are no orthogonality restrictions for the transmitted signals due to the MIMO signaling. Thus, MIMO MC-CDMA needs the processing on channel equalization and despreading. First we derive the BER performance of zero-forcing (ZF) detector and evaluate the performance of MMSE detector. In addition to this, for the lowcomplexity approach, we propose a low-complexity decision feedback detections for MIMO MC-CDMA systems. The proposed DF detection based on noise-prediction technique has the same BER performance as symbol-level detector with lower complexity when the system load is almost full. We also propose a partial minimum mean square error (MMSE)-...