Resource Management and Intelligent Design for Future Wireless Communication Networks

Nie, Dawei and Ni, Qiang and Pervaiz, Haris and Yu, Wenjuan (2026) Resource Management and Intelligent Design for Future Wireless Communication Networks. PhD thesis, Lancaster University.

Text (2024NiePhd)
2024NiePhd.pdf - Published Version
Restricted to Repository staff only until 1 March 2026.
Download (2MB)

Abstract

In recent years, data traffic has surged unprecedentedly, raising concerns over resource management as predictions forecast a continual exponential growth in devices and wireless connection demand. The increasing demand for high-security, sub-millisecond latency, ultra-dense networks, and ultra-low energy consumption has sparked concerns over resource scarcity. Effective strategies to address this challenge often entail optimizing resource consumption or managing the resource allocation of significant access traffic. Cognitive radios (CR) emerges as a promising solution to alleviate radio resource scarcity, enabling secondary users (SUs) to dynamically share licensed spectrum with primary users (PUs). Firstly, a novel cluster-based cooperative sensing-after-prediction scheme is proposed and optimized under a system accuracy requirement and a energy consumption constraint. The challenging integer programming problems are solved first by relaxing the integer variable. Then, two low-complexity search algorithms are proposed to achieve the global optimum. This work demonstrate that the total energy consumption and the number of users contributing to learning and sensing can be greatly reduced by applying our optimized clustered sensing-after-prediction scheme. Although the proposed algorithm significantly reduces energy consumption in CRNs, the inherent resource sharing nature renders the networks vulnerable to malicious attacks and reduce effective resource utilization. To ensure the security of our system and enhance the outcomes of our energy optimization efforts to increase effective resource utilization, minimizing the effects of malicious users (MUs) remains crucial. Two new types of MU effects model are proposed: normal negative effect and hidden negative effect, based on the behavior of malicious users in categorized groups. The effect models are utilized to formulate the two optimizations on decision fusion parameters then minimize the effect of MUs. The two optimizations base on MUs model yields minimal system error rates and effectively decreases the detection cycle for malicious user detection schemes, without significantly compromising decision accuracy. The outcomes of resource management for CR effectively meet the stringent requirements of both low energy consumption and high security performance. Finally, to fulfill the escalating demand for massive access in future wireless networks, it is crucial to enhance the current Random Access Channel (RACH) mechanism in 5G. A NOMA-enhanced 2-step RACH scheme that jointly leverages the benefits of the ACB, 2-step RACH, and NOMA-RA is proposed to reduce the latency and manage the resource allocation of massive access traffic. The latency performance and other theoretical trade-offs are analyzed by applying Markov chain model, while the optimal access probabilities and throughput of NOMA are derived for further optimization. To cope with the practical scenarios with constantly changing UEs traffic, the thesis proposes a Deep Contextual Multi-Armed Bandit (DCMAB) model that optimizes the NOMA throughput and dynamically adjust the barring rate to remain optimal latency based on the observable channel feedback, confirming the effectiveness of our proposed scheme. The outcomes of resource management for massive access effectively meet the stringent requirements of both low access latency and ultra-dense networks capability.