A primary video and a second instance of a reference video can be simultaneously transmitted to a system through at least a first node of a communication network, the second instance of the reference video configured to be compared to a first instance of the reference video stored by the system to generate at least one quality of experience (QoE) value that infers a perceptual quality of the primary video as received by the system. The QoE value can be received from the system. A determination can be made as to whether the QoE value is less than a threshold value. Responsive to determining that the QoE value is less than the threshold value, the perceptual quality of the primary video can be improved by re-routing transmission of the primary video to the system.