Dynamic Cluster-Based Resource Allocation Framework for Optimizing Video Streaming QoE Across Heterogeneous Network Conditions

Video streaming services face significant challenges in delivering consistent Quality of Experience (QoE) across users due to heterogeneous and dynamic network conditions. This paper proposes a novel framework that combines real-time clustering analysis of video streaming sessions with dynamic resource allocation to optimize QoE across diverse network environments. We introduce a modified STREAM clustering algorithm that efficiently groups streaming sessions based on key network metrics including bandwidth, round-trip time (RTT) and packet loss