| Multiple antenna techniques have been playing a leading role in modern wireless communications due to their advantage of the significant increases in capacity and reliability without additional channel bandwidth or transmit power. Spatial multiplexing, one of the prevailing techniques in the emerging wireless standards, such as 802.11n and 3GPP Long Term Evolution (LTE), can transmit multiple independent signals on the streams offered by the multiple antennas, providing high-speed data connections. In this dissertation, we consider singular value decomposition (SVD) based multiple beamforming where the transmitter as well as the receiver know the path gains on the multi-input multi-output (MIMO) wireless channel. In multiple beamforming, which employs a number of left and right singular vectors for the beamforming matrices, it has been known that diversity gain deteriorates as the number of transmitted signals increases. Bit-interleaved coded multiple beamforming (BICMB) was introduced to overcome the diversity loss of multiple beamforming. In this thesis, an extensive analysis of BICMB diversity gain is performed by calculating pairwise error probability (PEP), which depends on a subset of the eigenvalues of the Wishart matrix, thus requiring the marginal pdf of the eigenvalues for the calculation. We propose a methodology to calculate an upper bound to the marginal pdf of the eigenvalues, and apply it to the diversity analysis. Combined with the Singleton bound, the diversity analysis is also used to find the diversity bound of BICMB. Furthermore, the upper bound to the marginal pdf is applied to the diversity analysis of constellation precoded multiple beamforming (CPMB), which employs the constellation precoding technique in multiple beamforming. As a result, CPMB can achieve full spatial diversity offered by the multiple antennas with any code rate and any number of subchannels used. |
