Application Domains VoLTE



Transmitting voice packets with real-time requirements over a radio interface is a challenging task. Moving users, handovers, codec changes, varying radio conditions – just to mention the most obvious issues. LTE makes no difference.

But particularly speech transmission needs steadiness from beginning to end to deliver an excellent user experience. This is why a simple average quality summary per call does not help when assessing call quality.

VoLTE monitoring must not only deliver information on the control plane and S11 signaling, but also fixed-duration summaries of the speech quality.

Qrystal generates smart atoms (5-second summaries) for every RTP stream. In this way, the data can be used to determine for example

  • The quality of single calls or call directions
  • The quality of cells, eNBs, eNB clusters, or regions
  • The quantification of detected patterns
  • The impact of codec mode and bitrate change on call quality

Use cases of Voipfuture VoLTE monitoring

    1. Pre-deployment assessment
    2. RAN/EPC health check
    3. VoLTE troubleshooting
    4. Cell planning and optimization
    5. Analysis of bearer management
    6. Customer Care

Single Sided Calls

Favicon_Application Domains-86x86

Single Sided Calls

Everyone has experienced this

  • You can be heard, but you don’t hear incoming audio
  • Or vice versa
  • Or both

Whether it’s called single-sided call, one-way voice or silent call: For operation teams this is a hard problem.

To solve one-way voice you need to know the different reasons, for example

  • Missing RTP
  • Silent RTP
  • Too many RTP streams
  • Irregular RTP streams

Voipfuture Qrystal’s monitoring technology provides invaluable help. It is able

  • To automatically detect reasons for one-way voice
  • To help qualify the root-cause & quantify the impact

Monetizing MOS

Favicon_Application Domains-86x86

Monetizing MOS

Voice service monitoring is often tactical in nature. Tools produce data that is used by operations to work on customer tickets and to generally ensure that the network and service performs as designed. Any Return of Invest (ROI) mainly comes from savings due to higher operational efficiency.

But why not use the existing data in other areas to increase revenue and reduce cost? Qrystal and Qrystal Intelligence are successfully used by operations and performance optimization teams. On top, information provided by Qrystal is also used by CSPs to increase brand value and revenue by providing customers with access to their own service quality data.

Here’s why transparency pays off:

  • Differentiation (few CSPs are confident about their data)
  • Potential to sell data as a service
  • Fewer tickets (if data proves that problems originate in customer network)
  • Potential for selling troubleshooting services to customers
  • Customers love it

Of course, transparency can only work, if you can be confident that your data matches the user experience. The MOS and other media and signaling KPIs provided by Qrystal pass this test with ease.

Mean Opinion Score (MOS) - Calculation & Aggregation

Call Drop Detection

Favicon_Application Domains-86x86

Call Drops

Dropped calls are calls which were technically cut off before the call parties had finished their conversation. The dropped-call ratio (DCR) KPI is defined as the fraction of dropped calls over all established calls.

The problem: How to reliably detect call drops? What are the technical reasons for premature termination of calls? Qrystal eliminates the need for tedious configuration of proprietary SIP reason headers (RFC 3326), by learning from actual user behavior.

Users experiencing a call drop will often try to re-establish the call. The second call between two parties in short succession is referred to as ‘redial call’. These call pairs are analyzed to learn the set of SIP reason headers that have a high likelihood for redialing. In this way, call drop reasons are continuously analyzed and automatically learned by Qrystal.