User selection and resource allocation for secure multiuser MISO-OFDMA systems

The problem of user selection and resource allocation for the downlink of secure multiuser MISO-OFDMA systems overheard by multiple multi-antenna eavesdroppers is investigated. Based on zero forcing beamforming and artificial noise (AN), a physical-layer security model is proposed. Furthermore, the original problem is formulated as an AN aided secrecy rate maximisation problem, which maximises the sum of the secrecy rate by jointly designing user selection and the AN covariance. However, the formulated problem is non-convex and hard to optimise in general. It is shown that the problem can be handled by performing an exhaustive search in which a sequence of semi-definite relaxation programs are involved, leading to a tractable user selection and resource allocation method.