Managing telepresence service levels and users quality of experience (QoE)
Telepresence offers the best possible video conferencing and collaboration experience with a single or multiscreen high definition video and audio call in an environment that closely simulates a face to face meeting. Telepresence is now reducing travel requirements for many employees, saving money, time and of course carbon emissions. Initially marketed as a high end solution, the technology now allows telepresence end points to be installed at remote offices connected using lower bandwidth (typically over a dedicated T1/E1 circuit) and even over non QoS controlled networks. However, telepresence user expectations are understandably high so the requirement for a consistently good quality experience for both video and voice communications puts additional strain on the IT group responsible for managing service levels. The challenge becomes more difficult the more extended the deployment, especially when multiple network providers are involved.
It is important to realise that telepresence is designed to provide the best possible communications experience, and therefore the systems generally are intolerant of the typical IP network transmission issues that most other applications and services will tolerate. Many telepresence systems will take 'evasive' action to maintain the call, trying to overcome the network issues. In some cases the users will be warned that the system is trying to cope with an unreliable connection, and eventually the call may terminate automatically rather than give the users a poor experience. From an IT operations perspective this would be seen as a failure as the call quality was poor and the call had to be reconnected. The user may log a call with the service desk to report this failure however experience shows that most users don’t bother, they restart the call or resort to an audio only conference. If IT were even aware of this problem they generally have no information available to help them diagnose or even recognise this problem.
The Psytechnics solution can maintain Telepresence quality of experience
In order to effectively manage service levels IT must have a new approach, working proactively rather than just reactively, being able to identify and pre-empt any problems that may affect the user experience before the users start complaining.
With Experience Manager, IT operations has a real time dashboard view showing at a glance the performance of all voice and video applications. Telepresence application reporting shows the user's actual experience with IP voice and video. Psytechnics unique visibility into the application specific behavior of all calls presents the operational data in a clear and intuitive manner that enables operational and support teams to work proactively, with significantly greater levels of success; typically resolving user experience issues in a tenth of the time it would take using network management or QoS (quality of service) based tools, and doing it right the first time.
Psytechnics has led the way in perfecting voice and video assessment, performing subjective testing with thousands of people to validate the new generation of voice and video monitoring products. Psytechnics is a prime contributor to the seven most recent ITU-T standards for voice and video quality assessment. The proven accuracy of the Psytechnics monitoring algorithms means that Experience Manager reporting will precisely track the user's quality of experience.
Packet Monitoring vs. Video Performance Management
Packet based monitoring that looks only at the flow characteristics of voice and video network packets can't get the whole picture with respect to perceived voice and video quality. Experience Manager measures the quality of the individual media streams carried over the network in a manner consistent with the way humans perceive video distortions. This passive measurement process calculates the quality of a live stream by analyzing copies of the media packets at the monitoring points. Psytechnics software uses a detailed analysis of the packet transmission characteristics, codec in use, and the frame types impacted to predict the impact that packet transmission factors, such as packet loss and jitter, have on the perceived quality of the media stream when decoded by a particular edge device.
In order to monitor the quality of a voice or video stream accurately, it is necessary to account for the characteristics of the edge-devices (soft clients, video workstations, etc.) as the performance of these devices varies under different network and application conditions. This is due to the different coding, jitter buffer, and error concealment methods used. Psytechnics addresses this by accurately calibrating edge devices to understand their performance under the widest possible range of network and application conditions. Using this edge device specific calibration data, Experience Manager is able to more accurately detect and report problems that are affecting the user experience.
Psytechnics measurement technology also uniquely takes account of application performance such as bandwidth throttling, image resolution and frame-rate including changes that occur during a call. This coupled with Psytechnics network impact assessment, with its edge device calibration, provides a comprehensive assessment and diagnosis of end-user video QoE. The underlying video assessment technology is validated against extensive subjective test data and is supported by Psytechnics ITU-T J.247 (2008) active video assessment model.
To better aid troubleshooting when problems arise, video application information details such as the effect of compression, frame rate, bit rate, and video frame are reported. Video QoE information is reported as an easy to understand Mean Opinion Score (MOS-Video).
Psytechnics Experience Manager then compiles these data inputs to form the most complete predictive view of the user experience and uses that information to recommend alerts, alarms and ultimately corrective action such that:
| • | Support professionals can manage telepresence quality through real time predictive analysis of session-by-session experience |
| • | IT staff can measure experience impacts of specific MCU ports, session border controllers, codecs and endpoints |
| • | Detailed call visibility and analysis significantly reduces operations, support costs and resource requirements for troubleshooting real time IP video services |
| • | Highly accurate problem diagnosis enables accelerated corrective actions |
| • | Provides an objective view of IP video quality before-after state for project signoff when implementing new sites, new infrastructures or application upgrades |
| • | Delivers the rare yet extremely valuable ability to pinpoint sources of IP video quality faults, including external service providers |
IP video quality doesn't have to cost more
With Psytechnics Experience Manager, the many factors contributing to IP video quality, adjusted for the endpoint vendors' capabilities to compensate for network or codec effects are carefully and constantly monitored. Alarms are generated less frequently than typical QoS tools that lack equivalent endpoint awareness, and when an alarm is generated, it is for a valid reason so the IP video support staff can respond to IP video quality issues that really affect the user experience, and can take or recommend a far more prescriptive and accurate corrective action which accelerates time-to-repair.Reduced frequency of problems, reduced time-to-repair and a comprehensive end-to-end view of IP video quality are the keys to reducing support costs and improving the consistency of the IP video user experience.
