IoT-Enabled Secure Communication Protocols for Smart Grids

The growing need for secure communication for essential infrastructure, eases have intensified; smart grid functions are dependent increasingly on IoT devices. This research offers a framework based on the transformation of time series, domain specific feature extraction and supervised learning, to improve the communication security in smart grids. Continuous data of communication is partitioned into structured sequences by using time series transformation that can detect time irregularities in real time traffic. The key markers of secure and unsecure communication are extracted to characterise the domain specific features such as encryption techniques, packet latency and the transmission pattern. We implement the use of these models like random forest, Support Vector Machines to categorize the patterns of communication and identify the probable breaches. Experimental results indicate that the framework can achieve close to 96% accuracy on detecting secure protocols with relatively low number of false positive. To balance existing and upcoming cyber security requirements as related to communication security in the basic components of IoT enabled Smart Grids, this study advances an optimized scalable and resilient methodology.