Low-complexity lattice reduction architecture using interpolation-based QR decomposition for MIMO-OFDM systems

Lattice reduction (LR) technique is well known for its capability of improving MIMO detection performance. Nevertheless, the complexity increases significantly when LR is applied to all sub-carriers in the orthogonal-frequency division-multiplexing (OFDM) system. To address this issue, this research proposes an interpolation-based preprocessing architecture for LR-aided MIMO-OFDM system. The proposed architecture combines the interpolation-based QR decomposition and grouping architecture to decrease the iteration loops of LR by adopting lattice matrix of adjacent sub-carrier. The experiment compares the BER performance of the proposedmethod with those of other works in the channel of the 3GPP-LTE system. The experimental results show that the proposed LR architecture not only shortens latency but also reduce computational complexity and hardware cost for MIMO-OFDM systems.