Graduation Year

2020

Document Type

Dissertation

Degree

Ph.D.

Degree Name

Doctor of Philosophy (Ph.D.)

Degree Granting Department

Electrical Engineering

Major Professor

Richard D. Gitlin, Sc.D.

Co-Major Professor

Nasir Ghani, Ph.D.

Committee Member

Ismail Uysal, Ph.D.

Committee Member

Srinivas Katkoori, Ph.D.

Committee Member

Gabriel Arrobo, Ph.D.

Keywords

ALOHA-NOMA Protocol, Channel Capacity, M2M, MAC, SIC Receiver

Abstract

This dissertation is composed of two parts. The first presents several approaches to enhance the performance of 5G wireless systems by using NOMA (Non-Orthogonal Multiple Access) as the multiple access technique under different scenarios and performance metrics. The second investigates the performance of a wireless system network using a mobility model to evaluate the channel capacity taking into account motion. Both studies are directed towards improving the system performance in wireless communication systems.

In the first part, the optimum received power level for uplink power-domain NOMA with ideal Successive Interference Cancellation (SIC) reception is derived for any number of transmitters. The results show that the optimum received power level increases linearly (in dB) as the number of transmitters N are increased and the maximum required received signal-to-interference-plus-noise ratio (SINR) increases exponentially (or equivalently, linearly in dB) with the number of users N. An interesting observation is that the optimum power levels are very similar to that of the μ-law encoding used in the pulse code modulation (PCM) speech companders.

Next, a scalable, energy efficient, and high throughput medium access control (MAC) protocol, which is called the ALOHA-NOMA protocol, is proposed for Internet of Things (IoT) wireless 5G applications incorporating pure ALOHA with power domain NOMA. The simplicity of ALOHA and the superior throughput of NOMA and its ability to resolve collisions via use of a SIC receiver, makes ALOHA-NOMA an excellent candidate for a MAC protocol that can be utilized for a network of low complexity, low traffic IoT devices. The results show that the ALOHA-NOMA protocol significantly improves the throughput performance with respect to pure ALOHA, e.g., a SIC receiver that separates 5 signals can boost the throughput of classical ALOHA from 0.18 to 1.27 and with 100 active IoT devices the throughput is increased (at a greater than linear rate) to 40.

Following this, the bit error rate (BER) performance of a system with uplink power-domain NOMA and SIC reception using BPSK, QPSK, and 16-QAM modulations schemes in the presence of channel estimation errors is investigated. For each modulation level, two scenarios are considered: perfect channel estimation and a channel with estimation errors. The simulation results show that, for the two scenarios, the amount by which the BER of the SIC receiver increases as the modulation order is increased for a given noise level and the degree to which performance is degraded at high estimation error values. In addition to the BER performance study, for the same system model, a SIC detector degradation study in the presence of channel estimation errors, for the three modulations schemes, is presented. The study shows how the SIC degradation increases as a result of increasing the estimation errors and/or the modulation level. Somewhat surprisingly, for the three modulation schemes and for a small estimation error values, a linear relation is shown between the SIC degradation and the estimation error.

Finally, a channel capacity analysis for a random waypoint (RWP) mobility model in a wireless system network is introduced. The channel capacity for this model is derived for a Rayleigh fading channel with a maximum ratio combining (MRC) diversity receiver. The effect of the number of receiver branches on the channel capacity is analyzed and then the derived channel capacity is compared with the classic Additive White Gaussian Noise (AWGN) Shannon capacity and the static model Rayleigh fading channel capacity. The results show the level that the derived channel capacity increases as the number of MRC branches is increased. Also, the comparison shows that the AWGN channel capacity is greater than the RWP model channel capacity as it is not affected as severely by fading as the RWP mobility model. On the other hand, the RWP channel capacity is slightly larger than the static model Rayleigh fading channel capacity since severe fading will not affect the RWP model for as long a time duration as it affects the static Rayleigh model.

Share

COinS