Modeling, Debugging, and Tuning QoE Issues in Live Stream-Based Applications - A Case Study with VoIP

End-user acceptance criteria of live-stream based, interactive applications are different from traditional B2B or B2C applications. For example, if users sense disruptions in audio or video stream quality, they may quickly form a negative opinion. The Quality of Experience (QoE) in such live-stream applications is, thus, based on perception, and is open to subjective interpretations. QoE can be affected by hundreds of possible variables. QoE problems (QoE bugs), however, require an objective solution (fix in the product's code or tuning of product parameters). Understanding and debugging QoE bugs in such scenarios starts with designing relevant metrics and analysis tools. Thereafter, smart test-designs and strategies are required to gain insights into bottlenecks. This paper builds a discussion around these thoughts, and through a case study (sample QoE problem in a Voice over Internet Protocol (VoIP) application), collates some generic guidelines to investigate QoE bugs in live-streaming scenarios.