Performance Analysis of Matrix Inversion Algorithms for Massive MIMO Precoding under Rural and Urban Scenarios

Massive MIMO or Very Large Scale MIMO is a key technology in recent 5G and beyond 5G wireless standards to attain high speed, secure, error free communication and accessibility among trillions of users. All the above is possible with increase in complexity of the massive MIMO system. To improve massive MIMO system performance, efficient linear precoding algorithms are used at the downlink. The complexity of linear precoding algorithms lies on the hardware demanding large size matrix inversion module. In this paper, the performance of various approximations such as Neumann series, Conjugate Gradient (CG) and Preconditioned Conjugate Gradient (PCG) algorithms that simplify the matrix inversions of linear precoding algorithms are discussed. The simulations are carried out under various propagation scenarios like rural macro cell and urban micro cell.