Search

Energy infrastructure is of strategic importance to every nation as it plays a significant role in the development of a country. Smart grid, the new-generation power grid, focuses on duplex communication between suppliers and consumers. The heterogeneous nature of smart grid possesses different security threats, and integrating these components in the grid increases the possibility of vulnerabilities. The major challenge is to make sure that power availability is not affected when implementing security in the grid and hence a cyber-resilient smart grid is the need of the hour. Secure smart grid last-mile communication includes secure smart meters, smart homes and home area network. Internet of Things (IoT)-enabled smart homes are distributed throughout the smart grid home area network to facilitate the dynamic nature of grid. Smart meters communicate with all the entities involved in last-mile communication of the smart grid. A social Internet of Things based on trust model should be implemented for cyber resilience in smart homes for secure data transmission. The trust-based social IoT smart home architecture consists of machine learning-based predictive system for detection of cyber-attacks in smart homes. Machine learning approaches will be able to consider the computing capabilities of devices involved in smart homes and analyse the possible patterns of cyber-attack in the grid. This will help in early detection of attacks and threats in the smart grid. Machine learning-based approaches in smart homes can make the grid fault-tolerant and resilient which improves the security and efficiency of the smart grid.

Ever increasing need of electricity has paved the need for Smart Grids. Smart Meters, digitalized networks and fault tolerant systems are the basic infrastructure which supports Smart Grid. Security in Smart Grid has become a major concern in the present scenario. In this paper we have proposed security architecture at the last mile distribution in Home Area Networks. A Secure communication architecture has been modeled which focuses on secure data transmission between the Smart Meters at home and Central Gateway at the utility centre. Hybrid Encryption algorithms and Digital Signature has been used to provide data integrity. The strength of the model has been verified with the help of an attacker and the model is found to resist attacks. The Encryption time and Decryption time of the cyptostack is lower when compared with other encryption algorithms.

Strengthening of Smart Grid functionalities has become the need of the 21st Century. Security evolves to be the primary concern at the deployment level of Smart Grids. Cyber security threats and vulnerabilities in Smart grid Network needs to be addressed before the deployment of the Smart Grid. Our proposed intrusion detection scheme identifies anomalies in the Smart Grid traffic and detects attacks like flooding which causes Denial of Service in Smart Grid Networks. This paper applies k-Means algorithm for clustering of traffic data and outlier detection for the data transmitted between utility Centre and the Smart Homes. Performance of the algorithm has been compared with other clustering algorithms and the results were found to have higher percentage in anomaly detection.

Smart grid is a paradigm shift from the traditional Power grid which promises to make the electric grid both energy efficient and Fault tolerant. Trade-off between Energy savings and Security is a critical issue in Smart grid architecture. Smart grid architecture requires a high level secure data exchanges between sensors like Phasor Measurement Units and Advanced Metering Infrastructures like Smart Meters. In this study a Secure Smart grid Architecture model is proposed for the Smart grid network. Initially DamgardJurik encryption algorithm is applied on the data from the Phasor Measurement Units and a digital signature is then attached to the encrypted text to provide further authentication. The digitally signed data is collected in Data centre where it is decrypted. The proposed architecture has been implemented in both software and hardware. The effectiveness of the system is verified by introducing an intruder in hardware implementation.