Aiming at achieving better Quality of Service (QoS) provisioning in distributed wireless cooperative networks, a novel energy efficient jammer selection approach is proposed in this paper. We employ Secrecy Capacity (SC) to characterize the security of transmission. In order to accurately describe the time-varying characteristic, related channels are modeled as Finite-State Markov Channels (FSM Cs). The remaining energy of candidate node is considered in a similar way. Accordingly, jammer selection is modeled as Restless Multi-armed Bandit problem for the sake of energy balance and security. Primal-dual heuristic and priority-index heuristic are exploited to select the optimal jammer quickly and effectively. In the simulation, numerical evaluation of SC and networks lifetime results show that the proposed approach outperforms the existing ones. Aiming at achieving better Quality of Service (QoS) provisioning in distributed wireless cooperative networks, a novel energy efficient jammer selection approach is proposed in this paper. We employ Secrecy Capacity (SC) to characterize the security of transmission. In order to accurately describe the time-varying characteristic, related channels are modeled as Finite-State Markov Channels (FSM Cs). The remaining energy of candidate node is considered in a similar way. energy balance and security. Primal-dual heuristic and priority-index heuristic are exploited to select the optimal jammer quickly and effectively. In the simulation, numerical evaluation of SC and networks lifetime results show that the proposed approach outperforms the existing ones.