Iterative Interference Replica Subtraction Based Low-Complexity Precoding for Multiuser Massive MIMO Null-Space Expansion

This paper proposes a precoding weight that reduces computational cost and enhances interference suppression performance in time-variant multi-user massive MIMO (MU-mMIMO) environments. In MU-mMIMO, channel variations degrade the inter-user interference (IUI) suppression performance. Null Space Expansion (NSE) generates precoding weight that maintain nulls at the future channel state of interfering user terminals (UTs). On the other hand, mMIMO system requires high computational resources due to the large number of antennas. Iterative Interference Replica Subtraction (IIRS) eliminates the need for the matrix inversion calculation and provides a computationally efficient solution. The proposed precoding scheme integrates NSE and IIRS to address both challenges. We demonstrate significant improvements in computational efficiency and interference suppression under time-variant conditions through the computer simulation. The effectiveness of the proposed IIRS-NSE becomes particularly evident in scenarios with a small expansion factor and UTs' high mobility environment.