A new approach to alleviate the gain imbalance of parallel subchannels for MIMO systems

Eliminating the interferences among antennas is the first issue for effectively utilizing the spatial degree of freedoms of multiple-input multiple-output (MIMO) systems. Singular value decomposition (SVD) method plays an important role for interference elimination through a joint transceiver way, where the singular vectors are often used for linear unitary precoding (UP). However, directly applying SVD to the channel matrix suffers from subchannel-gain imbalance. In this paper, a new approach is proposed based on the lattice reduction (LR) method to alleviate such an imbalance problem. We apply SVD to the LLL-reduced channel matrix and obtain the new unitary precoding matrix. The additional complexity is negligible for long-block transmission. Numerical results show that the proposed scheme can outperform the conventional one for uncoded systems and coded systems.