Date of Award
5-1-2026
Document Type
Dissertation
Degree Name
Ph.D. in Engineering Science
First Advisor
Mustafa M. Matalgah
Second Advisor
Hamid Bahrami
Third Advisor
Lei Cao
School
University of Mississippi
Relational Format
dissertation/thesis
Abstract
This dissertation develops analytical and data-driven methodologies to address key challenges in emerging terrestrial and space communication systems, with emphasis on channel modeling, relay selection, and intelligent transmission strategies.
The first part establishes a unified analytical framework for decode-and-forward cooperative relay networks employing best relay selection (BRS) over generalized fading environments. By adopting the ?-? distribution as a flexible statistical model, closed-form expressions for outage probability, bit error probability (BEP), and ergodic capacity are derived using moment generating function (MGF)-based techniques. A cross-layer extension is further introduced by incorporating the impact of encryption, showing that error propagation effects impose intrinsic performance limits that cannot be alleviated by diversity alone.
The second part investigates terrestrial outdoor Terahertz (THz) communication under realistic propagation impairments. An optimization-based mixture-Gamma (MG) framework is first developed to approximate complex and composite fading channels with high accuracy. The proposed representation is validated using measured signal-to-noise ratio (SNR) data and enables tractable modeling of generalized fading conditions. Based on this framework, a comprehensive channel model is constructed that captures both deterministic effects, including free-space path loss and molecular absorption, and stochastic impairments such as small-scale fading, turbulence, and pointing errors. This formulation enables efficient and accurate evaluation of outage probability and achievable capacity.
The third part focuses on AI-driven communication strategies for low-power CubeSat systems operating in the Ka-band. A predictive adaptive modulation and coding (AMC) scheme is proposed, where machine learning models, including Long Short-Term Memory (LSTM), Gated Recurrent Units (GRU), and Temporal Convolutional Networks (TCN), are employed to forecast future SNR conditions. By enabling proactive link adaptation, the proposed approach improves average throughput and reduces outage probability compared to conventional reactive schemes. The results indicate that TCN-based predictors provide the most consistent performance in dynamic satellite communication scenarios.
This work provides a cohesive framework that combines advanced statistical modeling, cross-layer analysis, and machine learning techniques to improve the reliability and efficiency of next-generation communication systems across terrestrial, THz, and space-based platforms.
Recommended Citation
Alqodah, Mohammed Ali Mohammed, "Advanced Approaches to Emerging Terrestrial and Space Communication: Channel Modeling, Relay Selection, and Ai-Driven Transmission" (2026). Electronic Theses and Dissertations. 8794.
https://egrove.olemiss.edu/etd/8794