Parameters Measurement and Secrecy Diversity Analysis for Physical Layer Security in WSNs using Projection Pursuit Gaussian Process Regression

Abstract

Wireless sensor networks are specialized networks, geographically dispersed monitors that keep track of environmental external factors and conduct the collected information to a centralized opinion. The rapid growth of wireless sensor networks and its connected have pushed the saturation level of the communication. Moreover, the information passed is prone to the attacks and hence researchers have considered these as crucial factors in wireless sensor networks. Physical layer security is the one of the main approaches to ensure the secrecy of wireless sensor networks and has been attained with several encryption and signal processing approach. In our approach we propose a novel Projection Pursuit Gaussian Process regression (PPGPR) technique to analyze the diverse secrecy of physical layer in wireless sensor networks. The proposed approach is utilized to safeguard the sensitive data with secured physical layer. Simulations are performed and made an analogous study with state-of-art works to ensure the secrecy of the proposed work. The proposed work can use for the secured wireless communication with higher secrecy capacity than the existing approaches. The performance is analyzed with the parameters such as secrecy capacity, secrecy outage probability, packet delivery ratio, secrecy throughput, and energy consumption.