With the rapid evolution of 5G and beyond wireless networks, the demand for efficient, secure, and flexible spectrum management has become increasingly critical. This thesis investigates a blockchain-enabled framework for Dynamic Spectrum Access (DSA) and network slicing, aiming to improve spectrum utilization and enable trustworthy multi-stakeholder resource sharing. By integrating decentralized ledger technologies with advanced optimization and game-theoretic models, this work proposes novel solutions to address the challenges of spectrum scarcity, centralized control, and trust among participating entities. A major contribution of this work is the formulation and solution of a joint power and subcarrier allocation problem for Radio Access Network(RAN) slicing in a Multi-Tenant Multi-Service (MTMS) architecture. The proposed scheme considers heterogeneous network slices, including enhanced Mobile Broadband (eMBB), Ultra-Reliable Low-Latency Communications(URLLC), and massive Machine-Type Communications (mMTC), each with distinct Quality-of-Service (QoS) requirements. The optimization objective is to maximize System Spectral Efficiency (SSE) while meeting the performance constraints of each slice. Several algorithms were implemented and evaluated using MATLAB-based simulations, including Bounded Exhaustive Search (BES), Genetic Algorithm (GA), Surrogate Optimization (SO),and two Equal Resource Allocation schemes using GA. The GA-based Joint Power and Subcarrier Allocation (JPSA-GA) method achieved an SSE of 13.8 b/s/Hz, computed within 3 seconds, closely matching the 14 b/s/Hz upper bound attained by BES, which required over 4 hours of computation. Furthermore, the proposed algorithm achieves significant improvements in spectral efficiency (up to 6% higher than EPSO/ESPO and 20% over SO) while maintaining low computational complexity. This demonstrates that the proposed GA approach achieves a practical trade-off between performance and computational complexity. The thesis also develops a secure spectrum and resource trading framework using Ethereum-based public blockchain smart contracts. It introduces mechanisms for SLA enforcement, resource auctioning, compensation for SLA violations, and a reputation-based trust system for selecting reliable resource providers. Smart contracts are developed and tested using Remix Integrated Development Environment (IDE), an open-source environment tailored for building, simulating, and debugging smart contracts. Due to the inherent features of blockchain-based architectures, such as immutability, decentralization, and fairness, the proposed framework inherently enhances transparency, security, and reliability in the resource allocation process under diverse network conditions. Furthermore, a hybrid blockchain framework is proposed to enable efficient Dynamic Resource Sharing (DRS) between multiple Mobile Network Operators (MNOs). Hyperledger Fabric (HLF) is implemented on an Ubuntu Linux environment as a private blockchain for permissioned trading, while pricing and allocation strategies are optimized using a Multi-Leader Single-Follower (MLSF) Stackelberg game simulated in MATLAB. The buyer MNO’s strategy is solved analytically via Lagrangian duality, and the seller MNOs’ strategies are optimized numerically using GA, Pattern Search Algorithm(PSA), and Interior-Point Algorithm (IPA). Performance evaluation shows that the Raft consensus mechanism provides 57% lower latency compared to Kafka for the DRS process, demonstrating improved efficiency in transaction validation. The overall simulation results highlight the ability of the proposed framework to achieve pricing equilibrium and efficient resource utilization across multiple MNOs. To ensure SLA compliance in the multi-operator scenario, smart contracts developed in Solidity using the Hardhat Ethereum development environment were deployed and tested on multiple EVM-compatible public TestNets, including Goerli, Sepolia, Polygon Mumbai, and Shimmer EVM. The deployment and interaction were facilitated through Ethers.js and Node.js, while MetaMask was used for TestNet asset management. Among the evaluated networks, Shimmer EVM, an EVM-compatible TestNet for the IOTA Distributed Ledger Technology (DLT), demonstrated the lowest average transaction latency, approximately 87% lower than Goerli, and the highest throughput, confirming its suitability for time-sensitive SLA applications. This study assumes a static network model and does not explicitly account for user mobility or real-time channel variations, which are identified as directions for future work. Moreover, the applicability of blockchain-based mechanisms to latency-sensitive network services remains a limitation, as transaction and consensus delays may affect real-time responsiveness. Overall, the thesis establishes a robust and practical foundation for decentralized spectrum management and resource trading in future wireless networks, combining blockchain, optimization, and secure computing to address both technical and trust-related challenges.
| Date of Award | 20 Jan 2026 |
|---|
| Original language | English |
|---|
| Sponsors | Marie Skłodowska-Curie Actions MSCA |
|---|
| Supervisor | Pavlos Lazaridis (Main Supervisor) & Qasim Ahmed (Co-Supervisor) |
|---|