Healthy VoIP Nets�Part V�Quality of Experience
As transmitting video and TV become part the IP network's job, the basis for assessing quality becomes the user's experience.
When all of the dust in the laboratory settles, the testing methodologies that we described in the previous two tutorials (Part III, and Part IV) are focused in two directionsvoice channels and the networking infrastructure that provides the voice connectionmeasuring the impact that noise, echo, and other impairments have on the voice call. However, a new process, called Quality of Experience, endeavors to go beyond voice to include video and IPTV connectionsplus taking into account the effect that the end-user applications may have on the end-to-end connection.
And this topic is more than just the latest marketing buzzwordit was the subject of a three-day workshop sponsored by the International Telecommunications UnionTelecommunication Standard Sector's (ITU-T's) Study Group 12, the group chartered with dealing with transmission systems and media, telephone transmission quality, and the global information infrastructure.
For those of us that missed the trip to Geneva, Switzerland in June 2006, conference proceedings are posted on the ITU-T website http://www.itu.int/ITU-T/worksem/qos/200606/programme.html. According to the ITU-T, "this workshop brought together experts from all over the world who are involved in the definition and the delivery of end-to-end quality of experience (QoE) and quality of service (QoS). Several developments have been recently achieved on QoE and QoS, in particular related to NGN [next-generation network], but also in home networking, quality prediction, terminals, interoperability, and for new services." So let's dig in, and see how QoE differs from QoS.
Quality of Service deals with current network operating conditionsis there noise, crosstalk, echo, lost or dropped packets, or other maladies that are interfering with efficient packet delivery on the network, and/or do any of these aberrations cause the performance of the network to significantly degrade?
In many cases, an increase in the amount of network bandwidth, a more efficient allocation of the existing bandwidth within the network, or some type of prioritization or queuing mechanism can improve the QoS. Thus, QoS looks at performance from a network perspective, and may include functions that operate at the packet level to improve or otherwise manage the system.
In contrast, Quality of Experience is a measure of how well that network is satisfying the end user's requirements. QoE examines more than just the network infrastructure (or plumbing), and must also consider the end-to-end connection and applications that are currently running over that connection. More importantly, with multimedia networksthe direction where many VoIP implementations are headedadditional systems may be involved, including Internet video, 3G cellular, and IPTV networks, that both increase the network complexity, and add to user expectations.
For example, a videoconference with mediocre picture quality might raise higher user complaints than a similarly poor audio conference call, simply because the end users had much higher expectations (or possibly requirements) for the video-based event. In summary, QoE looks at performance from a user perspective, and with a focus on that user's applications.
Many new technologies start out on the service provider side of the equation, and then end up migrating to the enterprise side after some of the research investment has paid off. Such appears to be the case with QoE development, as much of the work is focused on the delivery of video in general, and IPTV in particular.
One organization that is working on quantifying QoE is the DSL Forum, a consortium of around 200 industry players that represent the telecommunications, equipment manufacturing, computing, networking, and service provider sectors. The organization was established in 1994, and since that time has developed a number of technical documents supporting DSL architecture, deployment and application support.
One of their recent publications is titled Triple-play Service Quality of Experience (QoE) Requirements, document TR-126, published in December 2006. This document recommends the minimum end-to-end QoE guidelines from an engineering perspective in order to effectively deliver triple play (Voice, Video, and Internet) services through a broadband infrastructure.
The applications addressed include voice, entertainment videoincluding broadcast and video on demandand data, including Internet access for web browsing and gaming. The document then provides details for each application in three categories: end-to-end QoE dimensions, QoE measurements, and QoE targets.
In addition, the document addresses the relationship between a performance-based QoS parameter (such as packet loss) and the resulting effect on the end user's QoE. In other words, given a QoS measurement, what is the predicted user QoE; or given a target QoE, what network performance (i.e., QoS) will be required to meet that objective. Engineering processes are then described that move the system under study toward the objective goal. And even though the document is slanted toward the service provider side of the house, it does not take too much creativity to move these concepts over to the enterprise side.
Our next tutorial will examine some of the specific VoIP transmission and quality parameters that the network manager needs to be keeping track of.
Copyright Acknowledgement: © 2007 DigiNet Corporation ®, All Rights Reserved
Mark A. Miller, P.E., is President of DigiNet Corporation®, a Denver-based consulting engineering firm. He is the author of many books on networking technologies, including Voice over IP Technologies, and Internet Technologies Handbook, both published by John Wiley & Sons.