In the field of physical layer security, fractional equivocation offers a broader perspective on secrecy, particularly regarding the confusion at the eavesdropper. Previous studies have utilized fractional equivocation to derive essential first-order metrics. However, deriving closed-form higher-order metrics based on fractional equivocation has been challenging due to numerical complexities. In this paper, we go beyond the first-order metrics and find a method to compute the higher metrics of fractional equivocation. We apply this method to simultaneous wireless information and power transfer (SWIPT) systems. Using the higher order moments, we introduce a new metric termed as amount of confusion loss (ACL). This metric assesses the level of confusion loss experienced by eavesdroppers, providing a more comprehensive characterization of their impact in real-world scenarios. Our analysis of ACL considers various power allocation scenarios to the energy harvesting receiver (EHR) and different signal-to-noise ratio (SNR) levels and illustrates the impact of these on eavesdropper's loss of confusion. This analysis reveals a significant disparity in the severity of confusion loss compared to the existing average leakage rate (ALR), particularly at low secrecy rates, which is not observed in previous studies. Furthermore, the influence of power allocation towards the EHR is distinctly reflected in the generalized secrecy outage probability (GSOP) and average fractional equivocation (AFE). Simulation results are performed to check the accuracy.
On the Secrecy of SWIPT Systems: An Amount of Confusion Loss Perspective
2024-06-24
629616 byte
Conference paper
Electronic Resource
English