In smart buildings and electric power maintenance and operation, it is necessary to obtain equipment or personnel location information in a timely fashion. There is a problem of low indoor positioning accuracy caused by non-line-of-sight transmission and the multipath effect. Thus a three-dimensional indoor localization scheme based on the combination of room division and 5G is proposed. First, the time difference of arrival (TDOA) and the arrival Angle (AOA) are used to integrate the positioning. Second, the specific positioning algorithm is integrated into the positioning architecture, and the location information of the corresponding moving target is quickly obtained based on an edge calculation. During TDOA positioning, an MEC positioning server performs channel estimation, and adds singular value for noise reduction on the basis of a segment orthogonal matching tracking (StOMP) algorithm. During AOA positioning, improved beam space transformation technology is used to construct dimension reduction. To ensure no loss of information in dimension reduction, the additional angle error is analyzed and processed and then the multiple signal classification (MUSIC) algorithm is applied for positioning. Finally, the 5GC core network server uses the Chan-Taylor algorithm for TDOA/AOA fusion localization. Simulation results demonstrate that the proposed positioning method can achieve accurate positioning of the moving target. This work is supported by the National Natural Science Foundation of China (No. 61971190).