Joint frequency-domain differential detection and equalization for DS-CDMA signal transmissions in a frequency-selective fading channel

It has been revealed that direct sequence code-division multiple access (DS-CDMA) can achieve a good bit-error rate (BER) performance, comparable to multicarrier CDMA (MC-CDMA), by using coherent frequency-domain equalization (FDE) instead of coherent Rake combining. However, coherent FDE requires accurate channel estimation. Pilot-assisted channel estimation is a practical solution, but its accuracy is sensitive to the Doppler spread. In this paper, a frequency-domain differential encoding and detection scheme is proposed for a DS-CDMA mobile radio. Joint frequency-domain differential detection and equalization (FDDDE) based on minimum mean-square error (MMSE) criterion is presented, where a simple decision feedback filter is used to provide a reliable reference signal for MMSE-FDDDE. Also presented is an approximate BER analysis. It is confirmed by both approximate BER analysis and computer simulation that MMSE-FDDDE provides good BER performance close to the coherent MMSE-FDE and shows high robustness against the Doppler spread; it outperforms coherent MMSE-FDE for large Doppler spreads. The proposed MMSE-FDDDE can also be applied to MC-CDMA. A performance comparison between uncoded DS- and MC-CDMA shows that DS-CDMA with MMSE-FDDDE achieves better BER performance than MC-CDMA with MMSE-FDDDE for small spreading factors.