Adaptive Power Allocation with Quality-of-Service Guarantees in Secure Wireless Communications

This paper investigates secure transmission and the corresponding resource allocation policy with statistical delay Quality of Service (QoS) guarantees in the presence of an eavesdropper. Our eavesdropper can be viewed as the wireless counterpart of Wyner's wiretapper. To measure the secrecy capacity with delay QoS guarantees of such a system, we propose a concept of secure effective capacity based on the effective capacity theory. Then an optimization problem aiming at maximizing the secure effective capacity subject to a given delay QoS constraint is formulated. Lagrangian dual-based approach is used to solve this problem. Theoretical results show that the optimal power allocation depends on the delay QoS requirements and the channel states of eavesdropper. Experimental results show that the proposed policy has a superior performance in improving the system throughput over other policies, which is coherent with our theoretical analysis.