Please use this identifier to cite or link to this item:
http://hdl.handle.net/10603/521978
Title: | Novel Secured Design strategies on a holistic smart grid model to maximize the efficiency of last mile communication |
Researcher: | Divya M Menon |
Guide(s): | Radhika N |
Keywords: | Computer Science Artificial Intelligence; Smart Grid; Deep learning; Machine Learning; power grid; y digital communication technologies Engineering and Technology |
University: | Amrita Vishwa Vidyapeetham University |
Completed Date: | 2023 |
Abstract: | Smart gird has evolved as a replacement to traditional gird, as the traditional power system fails to meet the demand of the suppliers in terms of generation, distribution and consumption. Sensors are deployed in the grid to detect fault and provide fault tolerance, thus making the grid smarter. Due to the critical role that the power system plays in our society, there is a common agreement that the power grid need to be secured to ensure continuous power. To envisage a secured architecture, our primary aim is to provide security to the data and the transmission grid. This is done by deploying a secure trust based communication architecture in Home Area network. a trust-based architecture is essential for smart homes in smart grids to ensure the reliability, security, and privacy of the system. By building trust among stakeholders, we can ensure the successful adoption and deployment of smart home in smart grid .For this the overall trust is calculated for each entity with in the smart home. A trust based MQTT protocol is developed and the performance parameters throughput, packet delivery ratio and transmission speed is proved to be higher than existing protocols. Research also showed that end to end delay ratio is also minimal. The work also develop a Machine Learning strategy to identify data attacks and network attacks between smart meter in Utility center and Home Area Network. Hybrid classifiers and single classifiers were analyzed and results show that Support Vector Machine with Gaussian Radial Basis kernel shows superior accuracy. Study showed that hybrid classifiers help to detect smart grid network attacks because they can leverage the strengths of multiple algorithms, improve the accuracy of detection, mitigate the impact of noisy and uncertain data and provide flexibility in model customization. Developing a machine learning strategy to identify attacks on smart meters and HANs is essential for protecting the security and reliability of utilities. This strategy can help to detect attacks early.. |
Pagination: | xi, 116 |
URI: | http://hdl.handle.net/10603/521978 |
Appears in Departments: | Department of Computer Science and Engineering (Amrita School of Engineering) |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
01_title.pdf | Attached File | 266.38 kB | Adobe PDF | View/Open |
02_preliminary page.pdf | 1.12 MB | Adobe PDF | View/Open | |
03_contents.pdf | 32.83 kB | Adobe PDF | View/Open | |
04_abstract.pdf | 27.49 kB | Adobe PDF | View/Open | |
05_chapter 1.pdf | 344.67 kB | Adobe PDF | View/Open | |
06_chapter 2.pdf | 152.47 kB | Adobe PDF | View/Open | |
07_chapter 3.pdf | 363.83 kB | Adobe PDF | View/Open | |
08_chapter 4.pdf | 1.07 MB | Adobe PDF | View/Open | |
09_chapter 5.pdf | 611.98 kB | Adobe PDF | View/Open | |
10_chapter 6.pdf | 640.15 kB | Adobe PDF | View/Open | |
11_chapter 7.pdf | 11.61 kB | Adobe PDF | View/Open | |
12_annexure.pdf | 114.3 kB | Adobe PDF | View/Open | |
80_recommendation.pdf | 277.54 kB | Adobe PDF | View/Open |
Items in Shodhganga are licensed under Creative Commons Licence Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0).
Altmetric Badge: