Secure Data Processing and Transmission Using Machine Learning Techniques

Dharman J

Abstract

The amount of data created by smart devices has drastically increased in recent years as a result of the quick technological improvements in those devices and their use in a variety of applications. As a result, the big data (BD) created by various devices may be too much for typical data analytics tools to handle. However, this exponential expansion of data provides numerous types of attackers with new chances to launch assaults by exploiting data analytics weaknesses (such as SQL injection, OS fingerprinting, malicious code execution, etc.). The linear scaling Rock Hyraxes swarm-based convolutional neural network addresses these issues. (LSRHS-CNN) centering 5G IoT data balancing and universally unique identification short input pseudo-random (UUDIS-ECC) hash-based secure data transfer (DT) are presented. The suggested technique comprises five phases: authentication, destination choice, validation, secure DT, and load balancing. The Length Nano ID (LNanoID) is initially produced in the authentication process of registration. If the LNanoID matches a previously saved LNanoID, the user may continue communicating. If the user has permission, the destination is chosen. After destination selection, the SiP hash function generates the hash code using the sender and recipient's public keys in the validation centre. Using the UUDIS-ECC algorithm, the IoT data is subsequently sent to the destination in a secure manner. During DT, the LSRHS-CNN algorithm is used to balance the 5G IoT data.