Bayesian CRLB for Joint AoA, AoD, and Channel Estimation Using UPA in Millimeter-wave Communications

We derive non-random and Bayesian Cramer-Rao lower bound (CRLB) for pilot-aided joint estimation of angle-of-arrival (AoA), angle-of-departure (AoD), and small scale channel gain in millimeter-wave (mmW) communication networks, in which transmitter and receiver are equipped with uniform planar antenna (UPA) arrays. Numerical results reveal that Bayesian CRLB decreases with an increase in Rice factor. Furthermore, using the derived Bayesian CRLB, we also observe that uniform design of beamforming and beamcombining codebooks, in which the angles are uniformly quantized, yield better performance than other types of codebook design.