Correlation-based phase synthesis approach for MIMO capacity prediction using antenna magnitude patterns

This paper concerns the estimation of MIMO capacity in a given propagation environment from the magnitude patterns of multiple antennas. In our previous contribution, we presented a simple approach of synthesizing phase on the magnitude patterns by taken into account the antenna separation. Although this approach works well even when the antenna separation is small and the signal correlation is high, the use of sophisticated matching networks can drastically reduce the correlation, which in turn affects its performance. Therefore, in this paper, we propose a small refinement to the approach by making use of the correlation between the magnitude patterns. Simulation results show that an average capacity estimation error of up to 0.5% is achieved for the proposed approach, as compared to up to 20.6% error
for existing approaches.