Blind Adaptive Subcarrier Combining Technique for MC-CDMA Receiver in Mobile Rayleigh Channel

A new subcarrier combining technique is proposed for MC-CDMA receiver in mobile Rayleigh fading channel. It exploits the structure formed by repeating spreading sequences of users on different subcarriers to simultaneously suppress multiple access interference (MAI) and provide implicit channel tracking without any knowledge of the channel amplitudes or training sequences. This is achieved by adaptively weighting each subcarrier in each symbol period by employing a simple gradient descent algorithm to meet the constant modulus (CM) criterion with judicious selection of step-size. Improved BER and user capacity performance are shown with similar complexity in order of O(N) compared with conventional maximum ratio combining and equal gain combining techniques even under high channel Doppler rates. Introduction: MC-CDMA is a well known multiple access technique which combines frequency-domain spreading and multicarrier modulation for effective operation in frequency selective multipath fading channel [1]. In conventional MCCDMA receivers, two subcarrier combining techniques namely, equal gain combining (EGC) and maximum ratio combining (MRC), are employed [1]. These techniques are, however, severely effected by MAI as the system loading increases, leading to poor detection performance. More advanced techniques incorporating multiuser detection, or for example, multistage RLS subcarrier-combining scheme recently proposed in [2] could be used to improve the performance. However, they are more computationally complex. A different approach is proposed in [3], that uses transmit power adaptations in time and frequency domain for MRC to reduce the degrading effects caused by weak subcarriers. However, it requires closed loop operation and very accurate channel estimation. In [4], MMSE filtering based scheme is proposed for operation in mobile Rayleigh channel for Multicarrier DS-CDMA. Although it shows much improved