Designing deep learning-enabled surveillance model with classified security levels for smart area networks

As the urban population grows and city infrastructures become more complex, the need for efficient and responsive security systems in smart buildings becomes increasingly crucial. Traditional security systems, which rely heavily on fixed surveillance cameras and sensors, face challenges in adapting to real-time situational changes and handling large volumes of data. To address these issues, this study leverages deep learning technology to enhance smart building security through two innovative surveillance adjustment algorithms for smart building with classified levels: the Centralized Surveillance Adjustment Algorithm and the Hierarchical Surveillance Adjustment Algorithm. These algorithms are designed to optimize the placement of security nodes and generate security barriers, ensuring comprehensive coverage and efficient resource allocation. The Centralized Surveillance Adjustment Algorithm monitor areas with the highest surveillance levels and adjusts surrounding regions’ surveillance accordingly, and allocates resources where they are most needed. The Hierarchical Surveillance Adjustment Algorithm classifies areas into different risk levels and adjusts surveillance hierarchically to prioritize high-risk areas. We developed a deep learning model to predict the required surveillance levels based on real-time data, facilitating dynamic and responsive security adjustments. We evaluate the performance of our proposed algorithms in a simulation environment. The results demonstrated that the Centralized Algorithm consistently outperforms the Hierarchical Algorithm in larger areas, providing superior coverage and adaptability.