Limited feedback Zero Forcing precoding scheme for MIMO-broadcast channels

This paper proposes a codebook based limited feedback zero-forcing(ZF) precoding scheme for multi-user multiple-input multiple-output (MIMO) downlink channels. In such system, each user needs to calculate a channel directional index (CDI) and a channel quality indicator (CQI), feedback them to the base station (BS). BS selects the set of active users, and calculates transmit beams through zero forcing method. In case of rank-one feedback, each user should combine the MIMO channel into a multiple-input single-output (MISO) channel through a receiver weight, then feedback CDI and CQI of this combined channel. Current existing schemes use the Hermitian of the left eigenvector associated to the maximum eigenvalue of the channel as the weight. The problem is that when there is multi-user interference, the feedback information is inaccurate. Focusing on this problem, we propose a new feedback scheme which uses a quasi minimum mean squared error (MMSE) weight to calculate CQI and CDI. By the simulations in 3GPP long term evaluation (LTE) environment, the proposed scheme is proved to achieve higher spectrum efficiency than the existing schemes.