Deep Temporal LSTM Regression Network (DTLR-Net) Model for Optimizing Quality of Video Streaming Quality in CDN-P2P Model

Due to the increasing popularity of live streaming, the need for more efficient and effective content delivery servers has become more prevalent. With the help of the latest generation of streaming models, such as the Content Delivery Network (CDN), it can reduce the burden on the servers. However, the quality of the video streaming process is still not always guaranteed due to the varying peer chunks and the capacity of the hosts. This paper proposes a hybrid approach that places the peers in a sub-layer, which takes into account the serviceability of the hosts. The members of the mesh topology are chosen based on the various factors that affect the quality of the stream. This study presents a novel approach that takes advantage of the potential of a Deep Temporal LSTM Regression Network (DTLR-Net) Model to predict the trajectories of mobile peers within network. Using deep learning tools such as Long Short-Term Memory (LSTM) networks, this novel technique addresses the intricate challenges posed by the dynamic movement patterns of mobile peers. This paper shows that the proposed hybrid structure is more effective than the traditional methods. It has a 25% increase in the peers' upload capacity and 19% decrease in the start-up delay.