Dual Visibility

Favicon-Te_Technology

Dual Visibility

No doubt, keeping track of registrations, call setups and the like is essential to VoIP service monitoring. But signaling performance says nothing about the in-call user experience. Call control and media transport is handled separately—by different layers, different protocols and sometimes even different network elements.

Consequently, VoIP monitoring must not stop at signaling protocols like SIP, but also consider media transport over RTP – only then you really understand your users’ view on your service quality. RTP Monitoring is very challenging, if just for the sheer number of packets that need to be analyzed. Yet, RTP Monitoring does offer the most valuable insights.

Complete VoIP analytics therefore needs to combine the views on the control and media plane to unfold the maximum benefit for VoIP service operations – this is what we call “Dual Visibility”.

Voipfuture has mastered both SIP and RTP Monitoring and our flagship product Qrystal is the only solution on the market to offer carrier-grade Dual Visibility.

Timeslicing

Favicon-Te_Technology

Slicing and Aggregation

Being able to group call data according to business-relevant criteria is key to many use cases. For example, enterprise customers of CSPs are increasingly demanding voice quality SLAs. With Qrystal you can obtain the data by defining a SIP trunk for each customer using the IP address of the SBC terminating the traffic of that customer.

A different example is given by a wholesale carrier who wants to report on call quality by origination and destination. This is done by introducing a numbering plan, which defines number prefixes, for example for countries and mobile networks. A-party and B-party numbers are matched to entries in the numbering plan. The link between calls and numbering plan entries can be used in a variety of ways, for example to report on the call quality of a business customer for calls to Singapore.

The configured trunk and numbering plan grouping is applied to each call and thus to all correlated streams. Throughout Qrystal, you may use these groups and aggregates to select a subset of calls or media streams to analyze or report on.

HD Voice

Favicon-Te_Technology

HD Voice

User experience is based on expectations – and they are rising. Many people making use of over-the-top (OTT) applications, such as Skype and Facetime, have enjoyed crystal clear speech quality for many years. Commercial voice service providers will have to adapt to the quickly changing expectations and embrace HD voice.

 

Modern HD voice codecs support numerous different modes and bandwidths. This would be of little importance to VoIP monitoring, if the in-call user experience would not be impacted by the specific codec parameters. The opposite is true. For example, the R-Factor of EVS WB bitrate modes ranges from 99 (7.2k) to 129 (24.4k), which translates to a MOSWB range of 3.91 to 4.5. According to ITU-T G.107 this leaves user satisfaction between “some dissatisfied” to “very satisfied”.

 

This means the actual user experience strongly depends on the codec and its parameters, which are not all under full control of a VoIP or VoLTE service provider. Quality can change on a packet by packet basis – put to the extreme, user satisfaction can change every 20ms.

 

Qrystal’s RTP monitoring technology analyzes every RTP packet – including the specific audio encoding – to accurately estimate the user experience for every 5 second timeslice.

Media KPIs

Favicon-Te_Technology

Media KPIs

Voipfuture developed a set of six media KPIs for next-generation voice services.

Voipfuture’s atomic quality units allow to analyze and understand media plane quality. These units are aggregated to different levels and provide different perspectives – each with their own focus:

  • Service Perspective: Focused view on the media quality of the service in terms of the customer experience
  • Network Perspective: Focus on the technical quality of service or network transport performance
  • Session Perspective: Focus on quality for entire streams indicating distribution of quality across call sessions
  • Volume Perspective: Focus on quality per minute (relying on time slices) indicating overall quality of the observed traffic
  • Mid-point Perspective: Focused view on the quality at specific points in the network, e.g. at network boundaries
  • Call Perspective: Focused view on the end-to-end-quality as perceived by the calling parties

Correlation

Favicon-Te_Technology

Correlation

Voipfuture Qrystal uses a powerful two-stage correlation mechanism to correlate all media and control plane data measured at different points in the network.

Qrystal Probes first apply a patented online correlation mechanism, which correlates signaling and media information while VoIP sessions and media streams are still in progress. The correlation mechanism is able to identify and track media streams in all phases of a session, including early media (ringtones, pre-call announcements) from multiple parties. Additionally, Qrystal recognizes media streams that have been modified on the IP layer to pass through firewalls and Network Address Translators (NAT). This correlation step also allows to associate xDR and RTP stream data with matching trunk and numbering plan definitions. This is subsequently used to create statistics and KPIs.

In the second stage, Qrystal correlates all xDRs and RTP stream data of a call measured at different points in the network to create call detail records (CDR) representing an end-to-end call view. All RTP streams and xDRs of the same call receive a CDR tag, so all of them are linked to their CDR.

GMR

Favicon-Te_Technology

Good Minute Ratio

Voipfuture developed a set of media KPIs for next-generation voice services.

Based on Voipfuture’s unique 5 sec timeslicing technology network performance data can be sliced and diced according to your needs.

As an example, here’s how to the Good Minute Ratio (GMR) is calculated:

  • A ‘good minute’ is a set of time slices representing one minute, where all time slices have a MOS above 4.0.
  • Good Minute Ratio (GMR)
    GMR = # good minutes / total number of minutes
  • GMR calculation in this example: GMR = 97/100 = 97%

The GMR can be calculated for any group of media streams, e.g. one direction of a call, all calls in the past 15 min to Bangladesh or for the incoming RTP from an interconnection partner in the past year.

It’s actually quite ingenious – all you have to do is count the number of Smart Atoms with a specific characteristic and you can get a real picture of the service quality you are providing your customers.

Understanding MOS

Favicon-Te_Technology

Understanding MOS

Voice quality is typically expressed in terms of the mean opinion score (MOS), which from its origins is a subjective rating of audio quality on a scale from 1 to 5. For many relevant use cases, such as customer experience management and SLA monitoring, it is not practical to ask test persons for their opinion. The original concept has therefore been extended to cover cases where MOS is not the result of an empirical study but the output of a computer-based analysis.

Today, there are many standards and recommendations describing algorithms for estimating different kinds of MOS. This includes objective MOS measurements obtained from active measurements (MOSLQO, MOSCQO) and estimated MOS values (MOSLQE, MOSCQE) from passive methods. In addition, MOS can be provided on different scales, depending on whether users can expect wideband high-quality audio or standard ISDN quality.

Voipfuture timeslicing technology calculates a MOS for each 5 second RTP segment offering a highly detailed assessment of the user experience.

Codec Detection

Favicon-Te_Technology

Codec Detection

For decades, the G.711 codec was considered the gold standard of telephony, even though it suffers from the limitations of narrowband signal transmission.

Modern HD codecs, such as AMR-WB, EVS and Opus, typically feature several modes with different bit rates, sampling rates and coding schemes.

Some of these modes yield quality that is actually worse than G.711 and thus cannot be considered “high definition”. For example, the user experience provided by the AMR-WB mode with 6.6 kbit/s is well below toll quality.

Although, seven out of nine AMR-WB modes deliver a quality better than G.711, the claim of using an HD codec does not guarantee excellent user experience. Such codec mode changes can occur on a packet-by-packet basis without any indication in the SIP signaling.

Most VoIP monitoring tools on the market are not capable of detecting the codec based on analysis of the RTP payload.

Qrystal is a notable exception and provides voice service providers worldwide with accurate information on user satisfaction.

Intelligent Alarming

Favicon-Te_Technology

Intelligent Alarming

Voipfuture developed a new intelligent alarming technology for Qrystal Intelligence, which only raises alarms that are significant, pointing to unusual events in a VoIP service. Conventional alarming compares current data to static thresholds. This approach is good for use cases such as SLA monitoring or policy control. Here, hard thresholds need to be met, because they are part of contractual agreements or enshrined in management goals.

Such hard thresholds do not help to detect anomalies. Voice service quality often depends on the network load, the route, destination and the hour of day and day of week. Setting static thresholds in such dynamic environments is almost impossible – either thresholds are too tight, raising too many alarms, or thresholds are too loose, making them effectively useless. In practice, VoIP service operation teams are often flooded with alarms preventing them from identifying the issues that truly matter.

Intelligent Alarming is designed to overcome the limitations of conventional alarming. Intelligent Alarming uses machine learning technology to predict future KPI values based on historical data. This adaptive thresholding detects events that are outside the norm and help operations teams focus on real incidents.

Signaling KPIs

Favicon-Te_Technology

Signaling KPIs

IETF’s RFC 6076 defines common key performance indicators (KPI) to assess a network’s SIP signaling performance, i.e. its general ability to set up and tear down calls, the ability of users’ to register, the corresponding delays, call durations, etc. Many of these KPIs adapt well-known telco metrics to the SIP-based world of voice over IP (VoIP). For example, the Session Establishment Effectiveness Ratio (SEER) is a SIP-version of the old Network Efficiency Ratio (NER) defined by the ITU.

Qrystal supports all RFC 6076 KPIs as they define the de-facto standard for VoIP signaling performance measurement. On top, Qrystal provides signaling metrics and KPIs based on unique SIP signaling indicators as well as LTE S11 metrics.