Link-Layer Rate of Multiple Access Technologies with Short-Packet Communications for uRLLC

Mission-critical applications such as autonomous vehicles, tactile Internet, and factory automation require seamless connectivity with stringent requirements of latency and reliability. These futuristic applications are supported with the service class of ultra reliable and low-latency communications (uRLLC). In this thesis, the performance of core enablers of the uRLLC, non-orthogonal multiple access (NOMA), and NOMA-random access (NOMA-RA) in conjunction with the short-packet communications regime is investigated. More specifically, the achievable effective capacity (EC) of two-user and multi-user NOMA and conditional throughput of the NOMA-RA with short-packet communications are derived. A closed-form expressions for the EC of two-user NOMA network in finite blocklength regime (short-packet communication) is derived, while considering transmissions over Rayleigh fading channels and adopting a practical path-loss model. While considering the multi-user NOMA network, the total EC of two-user NOMA subsets is derived, which shows that the NOMA set with users having distinct channel conditions achieve maximum aggregate EC. The comparison of link-layer rate of NOMA and orthogonal multiple access (OMA) shows that OMA with short-packet communications outperformed the NOMA at low SNR (20dB). However, at high SNR region (from 20dB to 40dB), the two-user NOMA performs much better than OMA. To further investigate the impact of the channel conditions on the link-layer rate of NOMA and OMA, the simulation results with generalized fading model, i.e., Nakagami-m are also presented. The NOMA-RA with short-packet communications is also regarded as the core enabler of uRLLC. How the NOMA-RA with short-packet communications access the link-layer resources is investigated in detail. The conditional throughput of NOMA-RA is derived and compared with the conventional multiple access scheme. It is clear that NOMA-RA with optimal access probability region (from 0.05 to 0.1) shows maximum performance. Finally, the thesis is concluded with future work, and impact of this research on the industrial practice are also highlighted.