The role of deep reinforcement learning in developing adaptive cybersecurity defenses for smart grid systems
Gespeichert in:
| Titel: | The role of deep reinforcement learning in developing adaptive cybersecurity defenses for smart grid systems |
|---|---|
| Autoren: | Hatoon S. AlSagri, Shahab Saquib Sohail, Shiju Sebastian |
| Quelle: | Journal of Information and Optimization Sciences. 45:2299-2307 |
| Verlagsinformationen: | Taru Publications, 2024. |
| Publikationsjahr: | 2024 |
| Beschreibung: | The integration of profound fortification learning procedures inside cybersecurity techniques has risen as a promising approach to invigorate the flexibility of savvy network frameworks against advancing dangers. Savvy lattice frameworks, with their complex organize of interconnected gadgets and basic foundation, show special challenges for conventional cybersecurity measures. In this setting, DRL offers energetic and versatile arrangement by leveraging its capacity to memorize from involvement and optimize decision-making in complex, energetic situations. This paper looks at the part of DRL in creating versatile cybersecurity guards custom-made particularly for shrewd network frameworks. By utilizing DRL calculations, such as profound Q-networks (DQN) and profound deterministic arrangement slope, shrewd lattice protections can independently adjust to changing risk scenes, distinguish peculiarities, and moderate assaults in real-time. Moreover, DRL empowers the creation of proactive resistance instruments that can expect potential dangers and preemptively alter security conventions. Through a comprehensive survey of existing writing and case thinks about, this paper highlights the adequacy of DRL in improving the vigor and flexibility of cybersecurity measures for keen lattice frameworks. Additionally, it investigates the challenges and openings related with the integration of DRL methods in viable cybersecurity systems, clearing the way for future inquire about and usage in securing basic framework against developing cyber dangers. |
| Publikationsart: | Article |
| ISSN: | 2169-0103 0252-2667 |
| DOI: | 10.47974/jios-1807 |
| Dokumentencode: | edsair.doi...........fb5c0efe1ef61cc6ee3b8c9e932526c0 |
| Datenbank: | OpenAIRE |
Nájsť tento článok vo Web of Science | Abstract: | The integration of profound fortification learning procedures inside cybersecurity techniques has risen as a promising approach to invigorate the flexibility of savvy network frameworks against advancing dangers. Savvy lattice frameworks, with their complex organize of interconnected gadgets and basic foundation, show special challenges for conventional cybersecurity measures. In this setting, DRL offers energetic and versatile arrangement by leveraging its capacity to memorize from involvement and optimize decision-making in complex, energetic situations. This paper looks at the part of DRL in creating versatile cybersecurity guards custom-made particularly for shrewd network frameworks. By utilizing DRL calculations, such as profound Q-networks (DQN) and profound deterministic arrangement slope, shrewd lattice protections can independently adjust to changing risk scenes, distinguish peculiarities, and moderate assaults in real-time. Moreover, DRL empowers the creation of proactive resistance instruments that can expect potential dangers and preemptively alter security conventions. Through a comprehensive survey of existing writing and case thinks about, this paper highlights the adequacy of DRL in improving the vigor and flexibility of cybersecurity measures for keen lattice frameworks. Additionally, it investigates the challenges and openings related with the integration of DRL methods in viable cybersecurity systems, clearing the way for future inquire about and usage in securing basic framework against developing cyber dangers. |
|---|---|
| ISSN: | 21690103 02522667 |
| DOI: | 10.47974/jios-1807 |