Massive MIMO and Full-duplex Relaying Systems

In this thesis, we study how massive multiple-input and multiple-output (MIMO) can be employed to mitigate loop-interference (LI), multi-user interference and noise in a full-duplex (FD) relaying system. For a FD relaying system with massive MIMO deployed at both source and destination, we investigate three FD relaying schemes: co-located, distributed cooperative, and distributed non-cooperative relaying. Asymptotic analysis shows that the three schemes can completely cancel multi-user interference and LI when the number of antennas at the source and destination grows without bound, in the case where the relay has a finite number of antennas. For the system with massive MIMO deployed at the FD relay, we propose a pilot protocol for LI channel minimum-mean-square-error estimation by exploiting the channel coherence time difference between static and moving transceivers. To maximize the end-to-end achievable rate, we design a novel power allocation scheme to adjust the transmit power of each link at the relay in order to equalize the achievable rate of the source-to-relay and relay-to-destination links. The analytical and numerical results show that the proposed pilot protocol and power allocation scheme jointly improve both spectral and energy efficiency significantly. To enable the use of low resolution analog-to-digital converters (ADCs) at relays for energy saving, we propose a novel iterative power allocation scheme to mitigate the resulting quantization noise via reducing the received LI power and numerically identify the optimum resolutions of ADCs for maximizing throughput and energy efficiency. For massive MIMO receivers employing one-bit ADCs, we propose three carrier frequency (CFO) offset estimation schemes for dual-pilot and multiple-pilot cases. The three schemes are developed under different scenarios: large but finite number of antennas at the receiver, infinite number of antennas at the receiver, and very small CFO, respectively.