VoLTE analysis based on network messages

Overview

VoLTE (Voice over Long-Term Evolution) is the EPC (Evolved Packet Core) and IMS-based (IP Multimedia Subsystem) high-speed wireless communication standard for 4G networks.

VoLTE issues can arise from the RAN (Radio Access Network), the backhaul/transport, the EPC or the IMS domain. Additionally, issues can also emerge from the telecommunications devices.

 

Required interfaces to conduct a VoLTE analysis

VoLTE call quality or audio quality issues are for example speech mutes, choppy audio, call drops, poor MOS (Mean Opinion Score) periods or other. To classify subscriber impacting events into the different telecommunications domains with causes assigned to them, various control plane messages need to be processed and additionally correlated with the user plane traffic. The following interfaces are required for a basic analysis with minimum tapping traces:

  • S1C – interface between eNB (evolved NodeB) and MME (Mobility Management Entity): It is strictly not necessary but gives additional insights about Radio Cause for issues experienced. This interface is using S1AP/NAS protocol for communication.
  • S1U – user plane interface between eNB and S-GW (Serving Gateway): Carries user traffic based on SIP/RTP/RTCP (Session Initiation Protocol/Real-Time Transport Protocol/Real-Time Control Protocol). User plane traffic on S1-U is warped into GTP-U protocol.
  • S11 – interface between MME and S-GW: It is needed to make correlation between U-plane and Mobile identities MSISDN/IMSI and to provide insight on Session handling and Handovers (Cell-ID). S11 communication is using GTPv2 protocol for communication.
  • S6a – interface between MME and HSS (Home Subscriber Server) for authentication, location & service information of the subscriber: It is necessary to get the NAS (Network Attached Storage) cipher/integrity keys (KASME). S6a is using Diameter protocol for communication.
  • Mw – interface between x-CSCFs (Call Session Control Function): In case the communication between UE (User Equipment) and P-CSCF is encrypted with IPSEC the Mw interface is needed to get the decryption key. Mx is using SIP protocol for communication.

The scope of tapping can be extended by adding traces for additional IMS interfaces as well as Gx and S10 interfaces.

 

Recommended approach to analyse network messages

An analysis of network messages allows to assign VoLTE issues to network domains and to derive root causes. One approach is to proceed in the sequence of error probability: RAN, EPC, IMS, Backhaul/Transport.

Step 1 - RAN issues

RAN issues can be recognized from eNB MDT L3 logs (Cells Trace) correlated with SIP/RTP issues. For example, speech Mute events can be recognized by: 

  • correlation with poor radio condition (RSRQ - Reference Signal Received Quality, RSRP - Reference Signal Received Power),
  • correlation with SINR (Signal Interference + Noise Ratio),
  • correlation with time frames of handovers and correlation S1AP/NAS events which indicate radio cause. 

A further step can be to drill down in worst cells with speech mutes. The result is that each impacted subscriber is classified with RAN root cause.

Step 2 - EPC protocol errors

Drill down in the EPC domain shows for the MME – MME (S10) communication causes for EPC protocol errors like Relocation Failure, Context Not Found or No Resources Available. With MME internal logs the main cause of speech mutes can be discovered.

Step 3 - IMS protocol errors

Further, drill down in the IMS protocol errors can show error causes like Media Gateway unequipped to detect requested Event, Insufficient Bandwidth etc. This shows that there is an issue with MRFP/MGW low capacity causing overload. This can be overcome by rebalancing MRFP/MGW load to other MGWs.

 

Detection of backhaul or transport issues

Backhaul or Transport issues can be indirectly classified if there are no RAN, EPC or IMS issues. Insights on backhaul transport for further investigation can be derived from problems on Up Link or Down Link, from combination with RTP/RTCP information etc. Confirmation of Backhaul or Transport issues can be complemented with active TWAMP (Two-Way Active Measurement Protocol) measurements.

As this analysis can be rather complex, automated processing and intuitive GUIs are required for fast and efficient operation.

 

Tool to run a VoLTE analysis 

Our Call Tracer tool offers easy insights of events across all protocols and offering automatic correlation per subscriber. Moreover, it enables you to find root causes 10x faster than other tools.

Call Tracer tool    Contact us

handcrafted by vividblack