Improving Video Streaming Quality in Congested Networks with In-Network Computing

Network congestion and consequent packet loss are challenges for video streaming. Although there is research on changing the video encoding to make them more resistant to packet loss, currently, network devices do not control the packet content to act better concerning this loss. This study proposes a new approach using in-network computing, a hardware module, and an algorithm to improve video streaming in congested networks to analyze the frame loss using different strategies. The results indicate a considerable decrease in frame loss with minimal impact on resource utilization and performance costs.