In the traditional cooperative jamming strategies,the interfering signals at the destination node are com- monly ignored,and many relays and a great amount of power are consumed.In order to reduce the consumption of total power and relays without degrading security capacity,group-sparse penalty is introduced in the optimization of security capacity,and a proximal gradient algorithm (PGA) is employed to solve the problem.Moreover,a co- operative jamming strategy on the basis of PGA is proposed.Simulated results show that the PGA-based algorithm helps obtain the security capacity effectively; and that,by adjusting the penalty factor,it can remarkably reduce the consumption of total delay power and delays only with small security capacity loss,reasonably allocate system resources and improve the service efficiency of relays.