Power Optimization Algorithms for Secure Communication via Cooperative Jamming Schemes: A Comparative Study

The ultimate objective of cooperative communication is to secure data transmitted from source to destination against eavesdropper attacks. The paper compares different power optimization algorithms for cooperative jamming schemes to improve the secrecy performance of a four-node Amplify and Forward (AF) relay network. The main focus of the work is on the investigation of Optimal Power Allocation (OPA) for maximizing the secrecy rate subject to a total power constraint using two cooperative jamming schemes, Source and Relay Based Jamming (SRBJ) and Source Based Jamming (SBJ). The secrecy performance of SBJ is evaluated for trusted and untrusted relaying scenarios. In the untrusted case, the situation in which an external eavesdropper and untrusted relay exist simultaneously is analyzed. The iterative algorithms such as the Nelder-Mead technique (N-M), the Broyden - Fletcher – Goldfarb - Shanno (BFGS) algorithm and the Conjugate-Gradient (CG) methods are used for power optimization and, consequently, for secrecy rate maximization. Both symmetric and asymmetric relay positions are subjected to the secrecy analysis. A comparison is conducted using the equal power Allocation (EPA) approach and the Exhaustive Search (ES) algorithm. The paper also studies the complexity of different algorithms and jamming methods based on the average number of iterations required for system convergence. It was found that the iterative algorithms provide better secrecy than the conventional methods. Experimental results reveal a trade-off between the iterative algorithms' convergence and complexity. The gradient-based BFGS and CG algorithms are less complex than the gradient-free N-M method. When assessing all jamming schemes, the N-M method is a good choice for convergence, whereas the BFGS is the best choice for lesser complexity.