Research On Cross-layer Based Resource Allocation With Statistical Delay QoS Guarantees For Secrecy System

Due to the broadcast nature of wireless medium, secure communication is always a crucial issue in wireless networks. Physical layer security, as an alternative to the traditional cryptographic encryption, can achieve security utilizing the physical layer nature from the information-theoretic perspective. Physical layer security has higher reliability as it doesn’t rely on the computing power.The real time multimedia applications will be the main developments for the next generation wireless networks. These services have higher demands to the delay Quality of Service (QoS) requirements. In addition, security is also important to these services. However, the physical layer security model isn’t suitable for the delay-sensitive multimedia applications, since Shannon theory places no restriction on the delay of the transmission. Hence, the first purpose of our work is finding an analysis model which is suitable for investigating the resource allocation problem with QoS guarantee in secrecy systems. Besides, in order to solve the contradiction between the limited wireless resource and the increasing QoS requirements, optimal resource allocation strategies which can realize both the high quality network services and energy-efficiency are proposed.Totally, the main works can be summarized as follows:(1) Based on the effective capacity theory, we innovatively propose a concept of secure effective capacity, and establish a cross-layer model. Combined with the convex optimization theory, this model can effectively solve the resource allocation problem with QoS requirements for the secrecy system.(2) Based on the secure effective capacity cross-layer model, the resource allocation strategy with QoS guarantee is investigated for the secrecy system which is constituted of a transmitter, a legitimate receiver and an eavesdropper. Two optimization problems, including secure effective capacity maximization and average power minimization are analyzed. Then, we solve the optimization problems via Lagrangian dual-based approach. Finally, a stochastic resource allocation algorithm which can operate without the knowledge of channel CDF is proposed.(3) As further application of secure effective capacity, the multi-user TDMA secrecy system is considered. We propose a joint power and time allocation scheme which can satisfy the security, the statistical QoS guarantee, the fairness among users and energy-efficiency simultaneously. We first construct an optimization problem which minimizes the utility function of average transmit power under the secure effective capacity constraint. Based on the Lagrangian dual technique and dual decomposition theory, the optimal strategy is derived via KKT conditions and solved by a subgradient iterative algorithm. Numerical results show the superior performance of our proposed strategy.