Optimizing Non-uniform Antenna Array For Massive MIMO Systems

As a key technology,massive multiple-input multiple-output(MIMO)has the advantages of deeply mining spatial resources,greatly improving system capacity and energy efficiency.In practical applications,Massive MIMO systems usually need to configure a large number of antennas.However,some studies have shown that when a large number of antennas are configured with a fixed physical size,the deployment of a conventional uniform planar array will affect the transmission quality of the signal.The specific performance is that in the shortdistance large-capacity transmission scenario,the distribution of channel eigenvalues is uneven,and multiple data streams cannot be transmitted.In addition,the placement angles of the array can also have an impact on the transmission performance,but the existing literatures have not theoretically explained the influence of the placement angle on the channel eigenvalues.In order to solve the above problems,this paper studies the optimal design scheme of nonuniform planar array(NUPA)based on point-to-point single-user transmission systems.In this paper,the influence of the non-aligned array system is taken into account,so that the proposed non-uniform array transmission system can further improve the signal transmission quality.Specifically,the antenna array designed in this scheme makes the distribution of channel eigenvalues more uniform,thereby increasing the channel capacity.The main contributions of this paper are:(1)In this paper,the geometric model of the array is established when the arrays are not aligned in a single-user point-to-point transmission system,and the influence of the angle on the channel eigenvalues is analyzed.The theoretical analytical relationship between channel eigenvalues and array position variables is derived by using the method of asymptotic analysis.Specifically,a deterministic line-of-sight(LOS)channel model is established under the transmission system,and the array geometry structure is theoretically analyzed.On the basis of the above model,combined with the Kronecker product characteristic,the channel eigenvalues are decomposed into horizontal component and vertical component,and the mathematical relationship between the channel eigenvalue component and the array element position variables is analyzed by the method of asymptotic analysis.In addition,combined with the matrix theory knowledge,we can draw a conclusion that when there is an array deflection angle in the transmission system the channel eigenvalue become smaller and the channel capacity will decrease.It can be further revealed that the obtained channel capacity value is the same as equivalent aperture,and the equivalent aperture is equal to the projection plane of the antenna array perpendicular to the transmission direction.(2)Based on the single-user point-to-point massive MIMO system with the arrays not aligned,this paper proposes a universal NUPA optimization design scheme,which can further improve the channel capacity.Specifically,based on the asymptotic results of the channel eigenvalues,the channel capacity expression including the position variables of the array elements is deduced.Then,combined with the derivation,the channel capacity maximization problem is further transformed into a mathematical problem of Vandermonde determinant maximization,and the numerical solution is obtained using the adaptive gradient descent method.The final simulation results show that the proposed NUPA structure has the higher system performance.That is to say,to achieve the same transmission performance,the proposed NUPA requires a smaller physical size.