Data security during information flow has been a difficult challenge in Internet of Things (IoT). To improve security status various conventional cryptographic methods such as public, private, and hybrid key generation had been utilized by various researchers. But these methods are associated with a lack of secure communication between two or more devices. In view of this, a lot of effort is being put into the physical layer of security. In this concept signal-based encryption and key generation are used as the major components in the security of data. In this research, we present a physical layer security technique based on wavelet transforms and Langrage’s interpolation. The wavelet based function precedes the sampling of Received Signal Strength (RSS) and convolution of the Langrage’s interpolation and generates a session shared key for communication. The size of the generated key is 128 bits and 1024 bits during the authentication process of sharing information. The proposed algorithm’s performance is evaluated through a simulation modeling environment. Different standard data bench for performance analysis is also used. The proposed technique is evaluated in two scenarios; Line of Sight (LoS) and Non-Line of Sight (NLoS). The proposed algorithm is contrasted with key generation techniques based on Lagrange and Skyglow (SKG).As compare to Skyglow and Lagrange’s algorithms, the proposed algorithm achieves a 3-5% improvement in overall efficiency for key generation validation.