| Signal space diversity, which achieves reliable communication in fasting Rayleigh fading channel by creating redundancy in signal space, is a power- and bandwidth-efficient diversity technique. However, the complexity of the optimal receiver grows exponentially with the diversity order we designed to achieve. In this work, we concatenate the signal space diversity scheme with a outer convolutional code at the transmitter, at the receiver we use iterative demodulation and decoding. By utilizing the soft output from the outer soft-input soft-output (SISO) decoder, we can do soft interference cancellation. We proposed two kinds of Gaussian approximations to calculate the soft output of the demodulator, one is the vector Gaussian approximation, the other is the scalar Gaussian approximation. The complexity of the vector Gaussian approximation grows cubically with the designed diversity order, while the complexity of the scalar Gaussian approximation grows linear with the designed diversity order. Both of these two method can exploit the signal space diversity very well. We also applied the two Gaussian approximation methods to do iterative demodulation and decoding for the concatenation of convolutional code and lattice-based space-time block codes. Their performances are compared with the linear MMSE method. Also, we analyzed the behavior of the vector Gaussian approximation method by using EXIT chart analysis.; When the convolutional code is concatenated with a modulator and a bit-interleaver is used in between, the mapping from bit sequences to the constellations affects the performance of the receiver very much. By a carefully designed mapping, we can achieve performance gain without adding additional complexity to the receiver or consuming any other resources. In our work, we considered the mapping from bit sequences to the space-time matrices. The mapping criterions are derived for the demodulator with perfect a priori and no a priori information. Optimized mappings are searched for some unitary space-time modulations schemes and non-coherent space-time modulation schemes.; Spatial diversity is more and more widely used today. However, to have spatial diversity, multiple antennas should be equipped at the transmitter and/or the receiver. This would increase the cost and the size of the transceiver in the mobile station. (Abstract shortened by UMI.)... |
