The modern world’s increasing reliance on automated systems for everyday tasks has resulted in a corresponding rise in power consumption. The demand is further augmented by increased sales of electric vehicles, smart cities, smart transportation, etc. This growing dependence underscores the critical necessity for a robust smart energy measurement and management system to ensure a continuous and efficient power supply. However, implementing such a system presents a set of challenges, particularly concerning the transparency, security, and trustworthiness of data storage and retrieval. Blockchain technology offers an innovative solution in the form of a distributed ledger, which guarantees secure and transparent transaction storage and retrieval. This research introduces a blockchain-based system, utilising Hyperledger Fabric and smart contracts, designed for the secure storage and retrieval of consumers’ energy consumption data. Finally, a user-friendly web portal was designed and developed using the node.js framework, offering an accessible and intuitive interface to monitor and manage energy consumption effectively.
24 January 2023 in Journal of Computer Virology and Hacking Techniques
Botnets continue to evolve despite many efforts by law enforcement agencies and security researchers. As a result, there is an increase in the number of cybercrimes. This has led to a greater research focus on botnet detection. Among the reasons for growth in botnet and cybercrimes despite greater research focus are that significant number of the proposed techniques are not reproducible (unavailability of source code), do not contain a detailed description for effective comparison, and the absence of a real world labeled dataset for effective comparison. There is a grave problem of the unavailability of the labeled real-world dataset for bot infection detection. This paper aims to create a public labeled real-world Domain Name System (DNS) dataset for bot infection detection. The dataset contains real world DNS traffic of benign and malicious hosts. The dataset containing 24 features is labeled to list infected Domain Generation Algorithms (DGA) hosts along with the botnet family name and the DGA domains used for C&C communication. A total of 7644 hosts were found infected with nine different botnets namely modpack, virut, necurs, conficker, ud3, suppobox, nymain, tofsee and pitou. Finally, a machine learning classifier is developed to distinguish DGA bots from normal hosts using these features with an accuracy of 99.59%.