LAUSR

The information security rate of wireless sensor network with multiple eavesdroppers and multiple relays has been a research hotspot in recent years. In the scene of channel state information varies slowly, prior state information of the channel can be used to get beamformer vector which maximize information security under the constrains of power. When the channel state information varies rapidly, above method is unsuitable. To tackle this problem, a stochastic beamforming algorithm is proposed in this paper. It introduces maximum time-varying information security as the objective function and takes the power expectation of the weight vector as the constraint, using variable relaxation techniques and the Charnes–Cooper transform to convert the non-convex optimization into a convex form. Furthermore, performance of proposed method using two kinds of time-varying weight vector, namely complex Gaussian distribution and elliptical distribution, is deeply evaluated. Finally, numerical results are presented to demonstrate the performance of our proposed schemes.