Nutaq’s Messaoud Ahmed-Ouameur and his team just released his latest IEEE paper that has been published in Nordic Circuits and Systems Conference (NORCAS): NORCHIP & International Symposium on System-on-Chip (SoC), 2015
Hardware impairments effects on Massive MIMO system’s capacity were investigated and seem to vanish asymptotically at the base station when equipped with a larger array. Surprisingly, the effect of the local oscillator (LO) phase noise does not strictly obey the same law as the additive noise when a common LO is utilized. Having a close look at large array base station utilizing OFDM signaling, one would expect that the sampling rate and carrier frequency offsets (SRO and CFO) to introduce similar effects as the phase noise due to the time (subcarriers and OFDM symbols indices) dependence of the channel vector’s phase terms. For practical reasons, we use simulation to investigate the effect of such offsets on massive MIMO utilizing low computational complexity precoding techniques such as conjugate beamforming (CBF) and zero forcing beamforming (ZFBF). A reduced computational complexity ZFBF utilizing polynomial expansion technique is considered to assess sensitivity to such offsets. The simulation results show that without accurate compensation or calibration methods, the effect of sampling rate and carrier frequency offsets is very dramatic on massive MIMO performance. Increasing the array size does not provide an economically viable solution to alleviate the negative effects of such offsets.
You have to be a IEEE Xplore member to access it.