Hybrid Nature Inspired Algorithm for Capacity and Secrecy Optimization in MIMO

Abstract

In this dissertation, a novel hybrid optimization framework is proposed for maximiz ing the capacity and spectral efficiency of Multiple-Input-Multiple-Output (MIMO) systems with spatially correlated antennas under Rayleigh fading channels. The sys tem model incorporates Doppler shift to account for mobility and applies a power allocation scheme optimized using a novel Hybrid Particle Swarm Optimization and Chameleon Swarm Algorithm (PSOCSA) that incorporates the combination of two adaptive swarm intelligence meta-heuristics; Particle Swarm Optimization (PSO) with global search proficiency and Chameleon Swarm Algorithm (CSA) which includes the adaptive exploration-exploitation mechanism. The Proposed algorithm maintains a population of particles that exploit the best-known solution while also escaping from local optima which in turn helps to speed up the convergence for a global optimum so lution. Comparative analysis demonstrates that the proposed PSOCSA algorithm sig nificantly outperforms other state-of-the-art algorithms in terms of both computational efficiency and capacity maximization across various MIMO configurations (4x4, 8x8, 16x16, 64x64). Subsequently, the problem is further extended to secrecy rate optimiza tion in MIMO wiretap systems by incorporating a MIMO-capable Eavesdropper in the MIMO network, where the goal becomes maximizing the secrecy capacity so that the legitimate receiver receives as much channel capacity as compared to the eavesdrop per. Simulation results show how the Hybrid PSOCSA consistently achieves supe rior performance compared to other standalone state-of-the-art algorithms for a variety of MIMO configurations — traditional (4x4 and 8x8) and Massive MIMO (16x16, 32x32, 64x64), as well as a realistic 5G setting with 128x128 antennas and a range of eavesdropper antenna arrays up to 64, providing maximized secrecy rates with reduced computational complexity and smaller standard deviations indicating faster conver gence and robustness. Moreover, the developed system model is designed with several practical factors, such as; antenna correlation, Doppler effect, interference power from neighboring cells, and imperfect Channel State Information (CSI), which represent the security challenges in real-world secure communication, and thus replicating compli cated and practical implications of modern wireless communication setups. Statistical tests like the Wilcoxon Rank-Sum Test and T-Test were applied to validate the perfor mance of the proposed hybrid algorithm. The numerical results exhibit the generality of the Hybrid PSOCSA to secure and enhance the diversity and robustness of next generation wireless systems in terms of security and overhead efficienc