Secure power allocation in cooperative NOMA networks using Harris Hawks Optimization

The problem of secure resource allocation in cooperative Non-Orthogonal Multiple Access (NOMA) network is difficult due to the secrecy-aware power control. The essential parameters like relay cooperation, inter-user interference, relay selection and the risk of eavesdropping in a highly non-convex optimization space optimize the resource allocation. In this paper, a single secrecy-conscious power-allocation scheme of relay-assisted downlink NOMA with passive eavesdropping, cooperative jamming, Decode-and-Forward (DF) and Amplify-and-Forward (AF) relaying are formulated. The objective is to maximize the effective secrecy throughput sum of the legitimate users through joint optimization of source transmit power and relay transmit power. The influence of optimization of source and relay transmit powers on jamming, codeword rate, and secrecy rate under practical communication constraints are also explored. Harris Hawks Optimization (HHO) algorithm is proposed as the numerical optimizer in this work. The proposed HHO combines the exploration and exploitation capabilities of the standard HHO with additional refinement strategies to improve convergence speed, solution stability, and optimization accuracy for secrecy-aware resource allocation. The analysis also bring out the role of the transmit power, the density of the relays, the strength of jamming, the rate of the codeword, and the rate of secrecy on the reliability-confidentiality trade-off. Besides performance gains, an enhancement of secrecy sum of about 6% in DF is obtained over AF when comparing the proposed HHO with Genetic Algorithm (GA). Also, around 17% improvement in computational run time is achieved with respect to GA. The performance of proposed HHO is validated using p-value, and 95% confidence interval statistical tests. The results indicate that the proposed HHO-based framework provides an effective and computationally manageable solution for secrecy-aware cooperative NOMA resource allocation problems.