Simulating Hybrid Deep learning mechanism for security enhancement during Blockchain based WSN authentication

Abstract

In wireless sensor networks, sensor nodes have constrained resources including processing speed, memory, and battery life. WSNs are susceptible to a broad range of vulnerabilities since they are often used in untrusted environments. Given the difficulties of securing a WSN, the reliability of the information it collects is also put into question. WSNs' authentication procedure permits checks on the legitimacy of both resources and data. Data in WSNs is protected against tampering thanks to authentication, which checks the data's provenance and allows only authorized changes. However, current authentication methods have certain security holes, such as those that may be exploited by ID spoofing attacks. When it comes to cyber security, blockchain is another example of a promising new technology. All transactions on the blockchain are cryptographically protected and cannot be altered once they have been made. The goal of this study is to one day use blockchain technology in wireless sensor networks (WSNs). In this research, we created a novel authentication procedure for WSNs that relies on blockchain technology. Users and a private blockchain were important in integrating sensor nodes and the blockchain into the study's system architecture. The study's data was subjected to a thorough examination for security. Deep learning model has been used to classify secure and unsecure record over blockchain based WSN. Classification of data stored on blockchain is made after performing filtering using optimizer.