| As lattice reduction (LR) technique can effectively improve the singularity of the channel matrix in MIMO wireless communications system, it is frequently applied as pretreatment of linear detection to improve the decoding performance. In this article, we innovatively study the applications of LR in MIMO system from two aspects, which are developing the Lattice-Reduction-Aided (LRA) detection algorithm with soft-information for the receivers and studying precoding technique for the transmit part, respectively.For the receivers, a sorted list based LRA detection algorithm with soft-information is proposed. This algorithm obtains the sorted lists of the transformed matrixes which corresponds to different channel matrixs, and stores these lists for detection query. When LRA detection is done to obtain survived vectors, these sorted list related to channel matrix can effectively reduce the number of detection nodes which needs to calculate and result in lower complexity.For the transmitters, this paper studies how to design the precoding matrix with LR. According to different duplex modes, we propose both LRA code-book-based precoding and non-code-book-based precoding mechanism. With LRA codebook-based precoding mechanism, the receivers quantify the channel matrix with lattice reduction quantization algorithm, and feed back the quantization bits to the transmitters for precoding. The forth chapter presents the construction of code-book for 2 antenna systems. Through extensive simulation results we show that this precoding mechanism not only guarantees full-diversity, but also avoids the complexity of choosing modulation and conding scheme repeatedly in current code-book-based precoding. For LR-based non-codebook precoding mechanism, the transmitters get the reduction matrix with LR algorithm and then the precoding matrix according to ZF and MMSE criteria. Simulation results show that this LRA-based non-code-book precoding method can dramatically improve the BER performance of the system. |
PDF Full Text Request