Distributed Jacobi eigen-beamforming for closed-loop MIMO systems

This work presents a novel eigen-beamforming technique suitable for tracking fading channels in multiple-input multiple-output (MIMO) systems employing low feedback rate, closed-loop feedback channels. The method uses a single unitary coordinate plane transformation to update the transmit antenna weights without requiring parameters that depend on fading rate, typically found in recursive methods. The computational complexity is fixed, in contrast with codebook solutions, whose operations count increases linearly with the codebook size. Simulations show that the new eigen-beamforming method reaches near optimal bit error probability performance in low mobility scenarios. Furthermore, by buffering the available feedback bits the proposed algorithm outperforms signed gradient methods for the same average feedback rate