An optimum security enabled System‐on‐Chip selection for Internet of Things applications using genetic algorithm modeling

Abstract

SummaryThe System‐on‐Chip (SoC) is the core component in embedded systems that push toward the evolution of the Internet of Things. The quick growth of embedded SoC devices has a substantial influence to overcome security issues in embedded system design. Since microcontroller‐based systems are security and safety‐critical, this article focuses on security aspects of state‐of‐the‐art microcontrollers through performing design space analysis of several crypto‐core algorithms. These crypto functions are analyzed and compared to identify and implement the appropriate algorithm for on‐chip security. The genetic algorithm (GA) optimization model is utilized for the optimal selection of secured SoC from the commercial devices that are stored in the database. In this work, the fitness function is proposed with various levels such as performance‐centric, security enabled, low operating power, and application‐specific SoC selection. We also validate the GA results by implementing the security benchmarks on resultant devices by analyzing the execution time and performance. The accuracy of the GA‐based optimum SoC selection is estimated using confusion matrix method that produces the accuracy of .