Scattering source localization method based on distribution matching in massive mimo

Abstract

In a large-scale MIMO system, a scattering information source locating method based on probability distribution matching comprises the following operation steps: (1) a signal preprocessing stage, namely an average autocorrelation matrix of a received user signal is computed, an antenna array response vector is multiplied by the average autocorrelation matrix, the vector representation of a probability density function of overlapped user direction of arrival (DOA) is obtained; (2) an iteration estimation stage, namely the obtained vector is normalized, the position information of users is obtained by matching of the probability density function, and then estimated signals are deleted from the normalized vector. After multi-time loop iteration in the process, the position information of all the users can be obtained. The approximate orthogonality of the antenna response vector of the large-scale MIMO system is used for converting the average autocorrelation matrix of the received signal to the vector, and a probability distribution matching method is used for obtaining the position information of the users. Under the premise that estimation accuracy of the user position information is improved, complexity of estimation can be lowered.

Description

technical field [0001] The invention relates to a method for locating scattered signal sources in massive MIMO, specifically, a method for estimating the center DOA and angle extension of scattered signal sources, and belongs to the technical field of wireless communication. Background technique [0002] With the continuous growth of data services, the number of existing antennas has been difficult to meet the system's requirements for spectrum efficiency. By increasing the number of antennas in the base station, the diversity gain and multiplexing gain of the system can be greatly improved, which forms a massive MIMO system. In a massive multi-user MIMO system, the base station is equipped with hundreds or thousands of antennas, and each user is only equipped with one antenna. Time division duplexing is adopted, and all signal processing, including channel estimation, signal detection and signal precoding, is performed in Base station processing, reducing the calculation b...

AI Technical Summary

Problems solved by technology

The search scope of these two methods only includes the central DOA and angle extension of a single user, and the complexity is much lower than ML and COMET, but each search still includes matrix calculations, so the complexity is still relatively high

[0004] In addition, the existing estimation methods all assume that the angular expansion of each user is small, and then use Taylor expansion to obtain an approximate expression of the signal covariance matrix to estimate the position information. However, this approximation is under the condition that the angular expansion is large There is a large error under , which will cause the performance of existing estimation methods to degrade with the increase of angle expansion

Images