A Fusion Approach of Edge-Cloud Computing for Optimizing Power Usage in Surveillance via Advanced Reinforcement Learning
Jagendra Singh, Neha Grag, Vinita Nagda, Deepti Shrimal, Hemant Sahu, Priya Kumawat · 2024
This paper introduces an advanced surveillance system aimed at tackling prevalent issues like limited data storage space, delays in event detection, accurate situation prediction, and high power consumption. To address these challenges, the paper proposes the use of Deep Reinforcement Learning (DRL) technology, leveraging its capabilities to enhance system efficiency. The architectural framework outlined in the study primarily focuses on optimizing energy utilization and video processing by efficiently leveraging both edge and cloud computing resources. The first and most important feature of this study is a thorough investigation of energy consumption levels specific to the specified area. Valuable insights are drawn from interactions with the systems, which play an important role in developing machine learning models capable of detecting optimal lines of activity based on acquired information about the patterns. In addition, our research focuses heavily on evaluating how well proposed solutions reduce latency, improve event accuracy, and minimize power consumption costs. It was analyzed and tested in detail to ensure that the system, suggested for the resolution of problems, meets and exceeds performance standards. This all-inclusive model introduces the study as a full-fledged probe into state-of-the-art technological advancements aimed at boosting surveillance efficiency.