In our previous work, we compared the theoretical bit error rates of multi-carrier direct sequence code division multiple access (MC-DS-CDMA) and multi-carrier code division multiple access (MC-CDMA). To ensure a fair comparison, we constrained both schemes to the same bandwidth, information rate, and energy-per-bit, and these constraints resulted in a possible trade-off between diversity gain and channel estimation errors between the two schemes. While only a single-input single-output (SISO) system was considered in our previous work, in this paper, we extend the comparison to a multiple-input multiple-output (MIMO) system which employs Alamouti space-time block coding at each sub-carrier frequency to achieve spatial diversity. We consider only those cases where MC-CDMA has higher frequency diversity than MC-DS-CDMA. Since increases in diversity yield diminishing gains, we conclude that the addition of spatial diversity to this multi-carrier comparison benefits MC-DS-CDMA more than MC-CDMA. To determine whether these gains for MC-DS-CDMA are enough to offset the difference in frequency diversity between the two schemes, we derive closed-form expressions for the bit error probabilities of both schemes, and we compare the MIMO results against those of the SISO system for different information rates, number of users, and number of pilot symbols per channel estimate.

This paper investigates the trade-off between diversity and channel estimation errors for asynchronous multi-carrier CDMA (MC-CDMA) and multicarrier direct-sequence CDMA (MC-DS-CDMA). We construct a side-by-side comparison between the two systems by assuming equal operating and channel conditions. Both systems perform pilot-symbol- aided channel estimation and maximal ratio combining at the receiver, where the weighting coefficients are determined by the channel estimates at each sub-carrier. Closed-form expressions for the probability of error of both systems are derived under two separate channel scenarios. These two scenarios assume a frequency-selective Rayleigh fading channel, where the coherence bandwidth is equal to either the bandwidth of a MC-DS-CDMA sub-band or the bandwidth of a MC-CDMA sub-band. Finally, we compare the bit error rates of the two systems for different information rates, number of users, and number of pilot symbols used in each channel estimate.