Linear dispersion codes design based on Grassmannian subspace packing

A new method to design codebook is proposed for Linear Dispersion Code (LDC) systems with limited feedback The extended codebook design method is introduced based on Grassmannian subspace packing and a mapping function, which is derived from the relationship between SVDs of the classical multiplexing systems and the LDC systems. Our proposed codebook can guarantee the maximum min-distance of the finite set. Simulation results show that, for 2×2 MIMO systems, if the feedback bit number is 4, the proposed codebook shows the similar performance to the optimal precoding codebook and has substantial performance advantage against the random search codebook.