论文部分内容阅读
Massive multiple-input multiple-output(MIMO) is a key technology for the 5th generation(5G)of wireless communication systems. The traditional plane wave channel model(PWM) is often not suitable for the large antenna structure, and in certain cases should be replaced by the more accurate spherical wave model(SWM). By using the spherical wave characterization method, this paper investigates the channel capacity performance of a linear massive MIMO system in line-of-sight(LOS) scenarios. Two types of access settings, the point to point(PTP) system and multi-user(MU) system, are considered. In the PTP setting, a geometrical optimization is performed to obtain configurations that are able to generate a full rank channel matrix for a linear massive MIMO system, which yields full spatial diversity even in LOS scenarios. Compared with the approximate and commonly applied rank-1 PWM, this is very useful for fixed wireless access and radio relay systems requiring high throughput. For the MU case, we compare the eigenvalue distributions of the LOS channels using the plane wave and spherical wave characterization method, and sum rate results are obtained by Monte Carlo simulations. The results show that MU systems using the more realistic and accurate SWM can achieve a higher sum rate than results from the PWM. This is beneficial and informative when designing massive MIMO wireless networks.