Deep Learning Approach for Cost and Storage Optimization of Video Streaming in Cloud Environments

Abstract

As the demand for video streaming continues to rise, cloud computing has become a vital infrastructure for delivering reliable and scalable services. However, cost management and storage allocation pose significant challenges in this context. This paper presents a novel application of deep learning techniques to optimize cost and storage utilization in cloud environments for video streaming. The proposed framework leverages a hybrid deep learning model that combines Convolutional Neural Networks (CNNs) and Long Short-Term Memory (LSTM) networks. CNNs are employed for pattern recognition, allowing the model to identify recurring patterns in resource usage. LSTM networks, on the other hand, specialize in sequence prediction and enable accurate forecasting of future resource demands. By analyzing historical video views, the model can make precise predictions about future resource needs. This enables dynamic resource scaling, allowing cloud infrastructure to be efficiently allocated in response to streaming demands. Moreover, the framework addresses storage optimization challenges by incorporating deep learning insights into data placement, replication, and retrieval strategies. The results demonstrate significant improvements in cost efficiency and storage utilization, with cost reductions ranging from 15% to 20% compared to state-of-the-art methods.