【摘要】 Cloud computing is always adopted to enhance the computing capability of mobile systems, especially when the mobile users prefer to use some computation intensive applications. Consequently, the distributed wireless relay infrastructure should be deployed to aid the traffic transmission. To further enhance the QoS provisioning goals of wireless cooperative network, this paper puts forward a multi-objective approach for distributed optimal relay selection, which takes Bit Error Rate (BER) and Secrecy Capacity (SC) into account simultaneously. Firstly, our proposal partitions the channel state into several levels according to the received signal-to-noise ratio (SNR) and describes the time-varying Rayleigh fading channel characteristics by using first order finite-state Markov model. Secondly, we model the relay selection as Restless Multi-armed Bandit optimal solution problem with respect to the channel state and the state transition probability. Finally, simulation results demonstrate the efficiency of the proposed approach which outperforms the existing ones.更多还原

【Abstract】 Cloud computing is always adopted to enhance the computing capability of mobile systems, especially when the mobile users prefer to use some computation intensive applications. Consequently, the distributed wireless relay infrastructure should be deployed to aid the traffic transmission. To further enhance the QoS provisioning goals of wireless cooperative network, this paper puts forward a multi-objective approach for distributed optimal relay selection, which takes Bit Error Rate (BER) and Secrecy Capacity (SC) into account simultaneously. Firstly, our proposal partitions the channel state into several levels according to the received signal-to-noise ratio (SNR) and describes the time-varying Rayleigh fading channel characteristics by using first order finite-state Markov model. Secondly, we model the relay selection as Restless Multi-armed Bandit optimal solution problem with respect to the channel state and the state transition probability. Finally, simulation results demonstrate the efficiency of the proposed approach which outperforms the existing ones.更多还原