Direct Positioning Method of Mixed Far-Field and Near-Field Based on 5G Massive MIMO System

With the increasingly complex urban environment and ultra-dense network in the fifth generation communication, the problem of near-field errors in angular positioning has become gradually prominent. In order to apply the azimuth detecting method in the future network, this paper presents a novel algorithm for the massive multi-input and multi-output system by dividing the base stations into near-field ones and far-field ones. First, the stationary beam-forming is utilized to fit the coarse positioning architecture we proposed, which determines the searching zone of direct positioning method. Then, the fourth-order cumulants are derived to fit the near-field model and an adaptive weight is calculated for the association of far-field base stations. Meanwhile, the error caused by the longitudinal array to the plane angle is taken into account and the linear weights are used to avoid multi-peak hopping errors. Finally, we build a searching formula to apply the direct positioning method in the coarse positioning zone. The developed algorithm which combines near-field and far-field can solve the problem of the increasing positioning error caused by 5G intensive cell with the pure angle of arrival method. Furthermore, our method has some resistance to multipath and non-line-of-sight, which improves the positioning performance of municipal hotspots to some extent.