Block diagonalization eigenvalue based beamforming precoding design for downlink capacity improvement in multiuser MIMO channel

In order to fully benefit from multiple-input multiple-output (MIMO) channels, multiuser techniques are adopted in the 3rd generation partner project (3GPP) long term evolution (LTE) downlink system. The core task of multiuser MIMO system is to suppress multiuser interference and enhance the signal to interference plus noise ratio (SINR). As a classical MIMO scheme, block diagonalization (BD) algorithm could eliminate the multiuser interference completely. Unfortunately, it also reduces the desired receive power and then has capacity loss especially at low signal to noise ratio (SNR). On the other extreme, eigenvalue based beamforming (EBB) maximizes the useful power of receive signals and takes no action to depress multiuser interference making it an interference limited system with high SNRs. Based on such observations, this paper proposes an appropriate combination of the above two schemes. In order to reduce the computing complexity of the proposed scheme, we present an alternative algorithm named as interference sorting (IS). Simulations over spatial channel model (SCM) are carried out with framework of 3GPP LTE technical specifications. It is verified that proposed scheme leads to a remarkable capacity improvement both with low and high SNRs. And the IS algorithm could reduce the complexity of system significantly with indistinguishable performance loss.