AI-Driven Load Balancing for Optimizing Cloud Resources in Social Network and Urban Event Data Analysis

Abstract

Cloud computing plays a pivotal role in managing large-scale data across distributed systems, particularly in the context of social network analysis (SNA) and urban event detection. Load balancing, a critical aspect of cloud infrastructure, ensures optimal resource allocation and efficient handling of tasks. This research presents a hybrid cloud load balancing system that integrates traditional algorithm-based methods with advanced AI techniques. The proposed system was evaluated using key performance metrics: Response Time, Throughput, and Resource Utilization.

Simulation results indicate that the proposed system outperforms traditional static and dynamic load balancing methods in all evaluated metrics. The response time of the proposed system (30 ms) was significantly lower than that of static (50 ms) and dynamic (40 ms) methods, indicating superior task processing efficiency. Additionally, the throughput of the proposed system (500 tasks/sec) surpassed both static (300 tasks/sec) and dynamic (400 tasks/sec) approaches, highlighting its capacity to handle large volumes of tasks. The proposed system also exhibited optimal resource utilization (85%), which outperformed the static (70%) and dynamic (80%) methods, ensuring efficient allocation of cloud resources.

The integration of AI-driven load balancing mechanisms further enhanced the system's adaptability to dynamic workloads, demonstrating the potential of combining algorithmic and AI-based strategies for cloud computing. The results underscore the proposed system's efficacy in optimizing cloud resources and improving the performance of cloud infrastructures handling complex workloads, such as social network data and urban emergency event monitoring.