Due to the asymptotically orthogonal channel, minimum mean square error detection algorithm is near-optimal for uplink massive MIMO systems, but it involves matrix
inversion with high complexity. This paper proposes a highparallelism detection algorithm in an iterative way to avoid the complicated matrix inversion. The parallelism level is analyzed and convergence is proved in detail. The proposed algorithm can
be implemented in a high level, which is equal to the max number of received data streams. The complexity can be reduced by one order of magnitude comparing with MMSE algorithm. Simulation results show that the proposed algorithm can closely
match the performance of the MMSE algorithm with few number of iterations. It also outperforms Neumann Series approximation algorithm in terms of block error rate (BLER) performance with same number of iterations.