Optimizing Video Streaming Services Using Edge Computing for Reduced Latency with Bandwidth Allocation Algorithms and Federated Learning

Due to the increasing popularity of online video streaming solutions, it is essential to develop new approaches which can significantly facilitate the broadcasting process to users and deal with the existing network limitations. The current work focuses on the enhancement of video streaming services by utilizing edge computing, bandwidth scheduling, and federated learning. Through the application of edge computing, data processing is distributed across multiple points, and latency reduction maximizes the delivery of videos on time. Maximal resource usage is maintained through the utilization of bandwidth allocation algorithms in a way that increases network traffic control during high usage. To continue, the service quality is promoted by the federated learning already used to analyse users' behaviour and predict content requests without violating the privacy of its customers. This collaborative, distributed ML framework allows edge server to learn streaming parameters from its own experience in order to allow the distributive optimisations of resources can be locally performed and the quality of service can be well maintained. The above-discussed model overcomes the latency, bandwidth and privacy issues associated with the traditional cloud-based models of streaming. Prototypes and extensive experiments prove that the proposed integrated system has low latency, high bandwidth utilization efficiency, and higher user satisfaction. This paper offers a practical, private, and economic solution that tackles the future requirements for video streaming in line with edge computing to aid the development of subsequent studies in edge-driven multimedia services.