LR algorithm based on MED for multi-input-multi-output LDs

Recently, lattice reduction (LR)-aided linear detectors (LDs) are shown to be very effective in multi-input–multi-output systems for low-complexity and small bit-error-rate (BER) performance. However, the LR algorithms in the previous LR-aided LDs mainly aim to improve the orthogonality of the channel matrix where only channel state information is used. In this study, the authors design a novel LR algorithm to enhance the BER performance of the LR-aided LDs. Unlike the previous LR algorithms, the proposed LR algorithm aims to decrease the modified Euclidean distance (MED) in the LR-aided LDs, whereas the MED utilises the received signal as well as the channel matrix. Note that the MED is directly related to the BER performance of the LR-aided LDs, thus decreasing the MED in the LR-aided LDs can help to reduce the BER of the LR-aided LDs. In the proposed LR algorithm, the partial column addition operation is used. The simulation results indicate that their LR-aided LDs exhibit smaller BER than the previous LR-aided LDs. Moreover, the computational complexity is also shown in the simulation results.