Resource Allocation Techniques for Non-Orthogonal Multiple Access Scheme for 5G and Beyond Wireless Networks

The exponential growth of wireless networks and the number of connected devices
as well as the emergence of new multimedia-based services have resulted in growing demands for high data-rate communications, and a spectrum crisis. Hence, new
approaches are required for better utilization of spectrum and to address the high data-
rate requirements in future wireless communication systems. Non-orthogonal multiple
access (NOMA) has been envisioned as a promising multiple access technique for 5G
and beyond wireless networks due to its potential to achieve high spectral efficiency
(SE) and energy efficiency (EE) as well as to provide massive connectivity in supporting the proliferation of Internet of Things. In NOMA, multiple users can share the same
wireless resources by applying superposition coding (SC) and power domain multi-
plexing at the transmitter and employing successive interference cancellation (SIC)
technique at the receiver for multi-user detection. NOMA outperforms conventional
orthogonal multiple access (OMA) by simultaneously sharing the available communication resources between all users via the power domain multiplexing which offers a
significant performance gain in terms of SE.
In this thesis, several resource allocation problems have been addressed in NOMA
based communication systems, in order to improve network performance in terms
of power consumption, fairness and EE. In particular, the NOMA scheme has been
studied in multiple-input-single-output transmissions where transmit beamformers are
designed to satisfy quality of service using convex optimization techniques. To incorporate the channel uncertainties in beamforming design, robust schemes are proposed
based on the worst-case design and the outage probabilistic-based design. Finally, the
EE is investigated for non-clustering and clustering NOMA schemes with imperfect
channel state information. To eliminate the interference between different clusters,
zero-forcing beamformers are employed at the base station. Theoretical analysis and
algorithmic solutions are derived and the performance of all these schemes has been
verified using simulation results.