Over-Sampled Beamspace Channel Estimation for Millimeter Wave Massive MIMO

Beamspace channel estimation for millimeter wave (mmWave) massive MIMO system is investigated. An identity matrix approximation (IA)-based beamspace channel estimation scheme is proposed, including the design of hybrid precoding and combining matrix as well as searching the largest entry of over- sampled beamspace receiving matrix. The design of hybrid combining and hybrid precoding is formulated as two optimization problems. By decoupling the design of analog combining and digital combining, closed-form solutions are obtained. Then an algorithm based on bisection search is proposed to search the largest entry of the over-sampled beamspace receiving matrix. Additionally, computational complexity is compared between the proposed scheme and the existing channel estimation schemes. Simulation results show that the proposed IA-based beamspace channel estimation scheme outperforms the existing schemes.