This document reports the activities related to the design of QoE Sensors. It is the first deliverable to be submitted within the WP5 activities. To ensure the coherency and fluidity of the coming WP5 deliverables, the current D5.2 details the design of Slicing QoE sensor while taking into account the QoE metrics defined within the three project use cases (Smart Grid, Smart City, eHealth). In this regards, the design aspects for a monitoring framework for raw and QoS and the QoE monitoring architecture were studied and elaborated.
QoE sensors are responsible for generating QoE metrics using the collected QoS values. This can be assured through a specific function that defines the relationship between QoS and QoE or using a predictive model for QoE based on classification/estimation that learns through training. Both options are depicted. QoE shall be understood to be a multi-dimensional concept which ranges over different aspects of quality and how users perceive it.
The code of the noisy neighbour model will be published on GitHub or gitlab when WP8 platform is available. The code details the training and prediction phases for anomalies detection model implementation with R language will be provided in GitHub.