Optimizing Cloud Analysis and Load Balancing for Efficient Social Network Services

Abstract

In practical applications, Social Network Analysis (SNA) plays a crucial role in understanding the structures that connect diverse individuals in interacting populations. This analytical approach has proven highly effective across various scientific fields. One area where it finds extensive use is in crowd sourcing, which involves gathering, integrating, and analyzing vast and diverse datasets generated from numerous sources in urban spaces, such as sensors, devices, vehicles, buildings, and human interactions. In today's world, urban emergency events can affect any country, community, or individual. As we rely more on cloud computing, a significant challenge arises in efficiently analyzing the data due to computational and communication imbalances among computers caused by the collective behavior of humanity. Traditional load balancing techniques require substantial effort to rebalance loads on the nodes and struggle to handle stragglers, making the resource allocation process even more challenging. To address these issues, it is essential to have a cloud computing simulator to test and validate proposed solutions before implementing them in real cloud infrastructures. Researchers have developed several cloud computing simulators for this purpose, which have been widely used within the research community. In our research, we have conducted a comprehensive survey of current cloud load balancing solutions. Based on our classification, these solutions can be grouped into three categories: General Algorithm-based, Architectural-based, and Artificial Intelligence-based load balancing mechanisms. Each category offers unique approaches to tackle the load balancing challenges in cloud computing. To evaluate these solutions effectively, we have identified suitable metrics and thoroughly analyzed their pros and cons. By doing so, we aim to contribute valuable insights into the effectiveness of different load balancing approaches in cloud computing scenarios. Ultimately, our research seeks to enhance the efficiency and performance of cloud infrastructures in handling complex and dynamic workloads.