Radio Resource Allocation in RIS-based Wireless Networks

Future wireless systems must deliver higher data rates and reliability while meeting strict power consumption limits and hardware costs. In recent years, researchers have begun using Reconfigurable Intelligent Surfaces (RIS) and Reconfigurable Holographic Surfaces (RHS) to control the way signals move through space, relying on dense panels made of simple reflective units that require little power. This thesis investigates the design, modelling, and optimization of such metasurface-aided communication systems with a focus on energy efficiency, spectral performance, and physical-layer security. The first part of the work analyses the role of active and nearly-passive RIS architectures under a global reflection constraint. Closed-form models for power consumption and achievable rate are derived, and new optimization frameworks are proposed for joint beamforming and reflection control. Results show that nearly-passive designs can approach, and in specific regimes surpass, the performance of active RISs when hardware losses are accurately accounted for. The analysis also clarifies how the power budget, amplifier efficiency, and the number of reflecting elements determine the optimal architecture. The second part addresses the Global Energy Efficiency (GEE) and Secrecy Energy Efficiency (SEE) of metasurface-assisted multi-user MIMO systems. Novel algorithms are developed by combining fractional programming, sequential convex approximation, and the Dinkelbach-type method. Both the transmitter and the RIS/RHS coefficients are jointly optimized under realistic power and hardware constraints. The proposed schemes guarantee convergence to stationary points and significantly improve energyper- bit and secrecy performance compared with conventional designs. The study also highlights the importance of explicitly optimizing for SEE rather than only maximizing the secrecy rate, as efficiency-driven formulations yield stable operation at moderate power levels without sacrificing security. The final part explores dual-metasurface configurations, in which a transmit-side RHS is combined with an environmental RIS. This setting enables fine control of both the transmitted and reflected wavefronts. A dedicated optimization framework is introduced to coordinate the two surfaces under the unit-modulus constraint, enabling joint adaptation of the transmit and reflecting metasurfaces. Simulation results confirm that a base station with a few radio-frequency chains, assisted by a dual-metasurface arrangement, can outperform a much larger fully-digital array in overall system efficiency. These findings establish dual-metasurface architectures as a viable path toward secure and sustainable sixth-generation (6G) wireless networks. All optimization algorithms were implemented in PYTHON and MATLAB using MOSEK and CVX/CVXPY toolboxes. Extensive numerical experiments validate the analytical developments and provide design guidelines for future metasurface-assisted systems. In summary, the thesis contributes new theoretical insight and practical tools for realizing energy-efficient, secure, and programmable wireless environments.