Space-Time Line Code for Massive MIMO and Multiuser Systems With Antenna Allocation

Abstract

This paper first investigates an M -by-2 massive multiple-input multiple-output (MIMO) system that transmits a single stream is investigated. For this system, we propose a space-time line code (STLC), which is a transmitting and combining (at a receiver) scheme that achieves full spatial diversity. For the STLC, two consecutive (time) information symbols are weighted as per channel gains (space), combined at each transmit antenna, and transmitted through the M transmit antennas for two consecutive symbol times. With two receive antennas, the STLC receiver simply combines the signals received in the two symbol times and achieves a diversity order of 2M (full diversity). We show that the proposed STLC asymptotically achieves the maximum (optimal) received signal-to-noise ratio as M increases with significantly reduced computational complexity compared with the optimal scheme. Because the proposed STLC receiver requires no or partial channel state information, it avoids the issue of massive MIMO channel estimation. Furthermore, the rigorous performance evaluation under spatially correlated and uncertain channel conditions reveals that the proposed STLC achieves comparable or better performance than the existing schemes, and the results verify that the proposed STLC scheme is a potential candidate for M -by-2 massive MIMO systems. Next, the transmit antenna allocation algorithms are devised for a multiuser STLC system. Each user achieves full diversity order from the corresponding MIMO channels after the antenna allocation. The signal-to-interference-plus-noise ratio (SINR) of each user is analyzed considering the multiuser interference and channel uncertainty, and its lower bound is derived. Using the SINR lower bound, greedy algorithms that allocate the transmit antennas are devised. Rigorous simulation demonstrates that multiuser STLC with the proposed antenna allocation is robust against channel uncertainty and can improve the average SINR, improving the quality of experience. Furthermore, it is observed that the proposed STLC with antenna allocation method achieves the best performance if M is sufficiently large. The results in this paper show that the STLC can be a potential candidate for an M -by-2 (multiuser) massive MIMO systems.