Optimal Reference Vector Selection Algorithm in Subspace Interference Alignment to Maximize System Throughput

Arbitrary reference vectors have been adopted in conventional subspace alignment as transmitting vectors at the transmitter side. Inter-cell interference among users can be eliminated using orthogonal vectors to the chosen reference vectors at the receiver side. However, in this case, the sum-rate varies using different reference vectors, even though the channel values remain constant, and vice versa. In this paper, we analyzed the variation of sum-rate corresponding to reference vectors. An optimal reference vector selection algorithm based on subspace interference alignment is proposed to increase multi-cell capacity. The proposed algorithm adopts reference vectors with small variance. Simulation results show that the proposed algorithms provide about 50 % higher sum-rate than does the conventional subspace interference alignment algorithm.