Subspace optimization-based interference alignment algorithms for MIMO interference channel

Two new interference alignment algorithms for MIMO interference channel (IC) are proposed. Both the algorithms are based on maximizing useful signal subspace and suppressing interference power (MUSI). The first algorithm is based on the convex optimization. Interference leakage is constrained and the useful signal subspace optimization function is built. Semidefinite programming relaxation and rank 1 decomposition techniques are used for searching the precoder and decoder. The second algorithm is optimized on the Grassmann manifold. Conjugate gradient method is used for searching on the Grassmann manifold. Alternating optimization method between precoder and decoder is used for both the algorithms. Simulation results show that both the proposed algorithms perform well at high SNR.