Efficient codeword search algorithm for limited feedback MIMO beamforming systems

Abstract

We present an efficient codeword search algorithm for limited feedback MIMO beamforming systems. Given a predefined codebook, identifying the best beamforming codeword may easily become inefficient in terms of computational complexity. To reduce the computational complexity during the codeword searching process, we propose a novel (off-line) preprocessing scheme, which specifies the feasible condition for efficient searching. By exploiting the triangle inequality principle for distance, we develop a promising kick-out condition to avoid computing the chordal distance in the presence of non-optimal codewords of the input vector. Numerical experimental results have shown that the computational burden of our proposed algorithm has been significantly reduced compared with the traditional exhaustive search.