Subjective Quality Analysis of Streaming Video on Campus Networks

Abstract

This paper presents an adaptation on analytical frameworks for the estimation of the subjective quality of streamed video. While past efforts have focused on ISP-level network analysis and quality estimation (usually for the purpose of resource allocation), this paper seeks to establish a framework for quickly quantifying both the network and environment factors that affect perceived streaming video quality. The framework developed investigates both the technical network aspects which affect streaming video (such as bitrate and packet loss, among others) as well as the nonnetworking aspects of video playback that affect perceived quality (such as screen size, playback device, and other environmental factors). This framework is then used to produce a real-world subjective video quality estimate for students living on the Princeton campus.