The manuscript entitled as QoE-driven dynamic management proposals for 3G VoIP services authored by Jose-Oscar Fajardoa, Fidel Liberal, Is-Haka Mkwawa, Lingfen Sun and Harilaos Koumaras has been accepted for publication by the Journal of Computer Communications, Elsevier.
In this paper the authors propose a combined approach where service level adaptation is considered first and, provided that no suitable parameter combination is capable of providing enough QoE, a change of network state will be suggested. In order to do so we analyse the performance of the end-to-end (e2e) performance metrics in this convergent scenario, the root causes of possible degradations and, finally, the combined effects of the different network segments and their impact on the user perceived QoE. We show the the map of best performing VoIP configurations for every state of the network segments. Furthermore, considering each of these configurations, we analyse the acceptability of the service or the convenience of trying to modify the network state. Finally, a lightweight implementation based on simple network state estimation and decision heurisitics is proposed and validated in terms of accuracy and responsiveness.