Power Allocation in Uplink NOMA-Aided Massive MIMO Systems
Nguyen, Khai The
MetadataShow full item record
In the development of the fifth-generation (5G) as well as the vision for the future generations of wireless communications networks, massive multiple-input multiple-output (MIMO) technology has played an increasingly important role as a key enabler to meet the growing demand for very high data throughput. By equipping base stations (BSs) with hundreds to thousands antennas, the massive MIMO technology is capable of simultaneously serving multiple users in the same time-frequency resources with simple linear signal processing in both the downlink (DL) and uplink (UL) transmissions. Thanks to the asymptotically orthogonal property of users' wireless channels, the simple linear signal processing can effectively mitigate inter-user interference and noise while boosting the desired signal's gain, and hence achieves high data throughput. In order to realize this orthogonal property in a practical system, one critical requirement in the massive MIMO technology is to have the instantaneous channel state information (CSI), which is acquired via channel estimation with pilot signaling. Unfortunately, the connection capability of a conventional massive MIMO system is strictly limited by the time resource spent for channel estimation. Attempting to serve more users beyond the limit may result in a phenomenon known as pilot contamination, which causes correlated interference, lowers signal gain and hence, severely degrades the system's performance. A natural question is ``Is it at all possible to serve more users beyond the limit of a conventional massive MIMO system?''. The main contribution of this thesis is to provide a promising solution by integrating the concept of nonorthogonal multiple access (NOMA) into a massive MIMO system. The key concept of NOMA is based on assigning each unit of orthogonal radio resources, such as frequency carriers, time slots or spreading codes, to more than one user and utilize a non-linear signal processing technique like successive interference cancellation (SIC) or dirty paper coding (DPC) to mitigate inter-user interference. In a massive MIMO system, pilot sequences are also orthogonal resources, which can be allocated with the NOMA approach. By sharing a pilot sequence to more than one user and utilizing the SIC technique, a massive MIMO system can serve more users with a fixed amount of time spent for channel estimation. However, as a consequence of pilot reuse, correlated interference becomes the main challenge that limits the spectral efficiency (SE) of a massive MIMO-NOMA system. To address this issue, this thesis focuses on how to mitigate correlated interference when combining NOMA into a massive MIMO system in order to accommodate a higher number of wireless users. In the first part, we consider the problem of SIC in a single-cell massive MIMO system in order to serve twice the number of users with the aid of time-offset pilots. With the proposed time-offset pilots, users are divided into two groups and the uplink pilots from one group are transmitted simultaneously with the uplink data of the other group, which allows the system to accommodate more users for a given number of pilots. Successive interference cancellation is developed to ease the effect of pilot contamination and enhance data detection. In the second part, the work is extended to a cell-free network, where there is no cell boundary and a user can be served by multiple base stations. The chapter focuses on the NOMA approach for sharing pilot sequences among users. Unlike the conventional cell-free massive MIMO-NOMA systems in which the UL signals from different access points are equally combined over the backhaul network, we first develop an optimal backhaul combining (OBC) method to maximize the UL signal-to-interference-plus-noise ratio (SINR). It is shown that, by using OBC, the correlated interference can be effectively mitigated if the number of users assigned to each pilot sequence is less than or equal to the number of base stations. As a result, the cell-free massive MIMO-NOMA system with OBC can enjoy unlimited performance when the number of antennas at each BS tends to infinity. Finally, we investigate the impact of imperfect SIC to a NOMA cell-free massive MIMO system. Unlike the majority of existing research works on performance evaluation of NOMA, which assume perfect channel state information and perfect data detection for SIC, we take into account the effect of practical (hence imperfect) SIC. We show that the received signal at the backhaul network of a cell-free massive MIMO-NOMA system can be effectively treated as a signal received over an additive white Gaussian noised (AWGN) channel. As a result, a discrete joint distribution between the interfering signal and its detected version can be analytically found, from which an adaptive SIC scheme is proposed to improve performance of interference cancellation.
DegreeMaster of Science (M.Sc.)
DepartmentElectrical and Computer Engineering
SupervisorNguyen, Ha H.
CommitteeBerschied, Brian; Mohamed, Ebrahim B; Zhang, Chris
Copyright DateSeptember 2020
Massive MIMO, NOMA, SIC, optimization