VoLTE troubleshooting: 10-minute field diagnostic method

VoLTE issues are uniquely challenging because they span three distinct layers: radio, IMS signaling, and media (codec/RTP). A voice quality complaint that manifests as “the call sounds choppy” could originate from poor RF conditions causing BLER on the QCI 1 bearer, an IMS routing misconfiguration dropping SIP packets, or a codec mismatch producing sub-optimal audio encoding. This guide presents a systematic 10-minute field diagnostic method that isolates the failure layer and identifies the root cause.

VoLTE architecture: the three layers

Before diagnosing VoLTE issues, engineers must understand the three interdependent layers that make a VoLTE call work.

Layer 1: Radio bearer (QCI 1)

VoLTE voice traffic is carried on a dedicated EPS bearer with QCI 1 (QoS Class Identifier 1), which provides:

• GBR (Guaranteed Bit Rate) — typically 40-80 kbps depending on codec
• Priority level 2 — highest priority for user plane traffic
• Packet delay budget: 100 ms — much stricter than default bearer (300 ms)
• Packet error loss rate: 10^-2 — one packet lost per hundred is the maximum tolerance

The QCI 1 bearer is established during call setup via a dedicated bearer activation procedure. If the bearer fails to establish, or if radio conditions cannot sustain the GBR, the voice call fails or degrades.

Layer 2: IMS signaling (SIP)

VoLTE call control uses SIP (Session Initiation Protocol) over the IMS (IP Multimedia Subsystem) infrastructure:

• P-CSCF (Proxy-Call Session Control Function) — first IMS contact point, typically co-located in the mobile network
• I-CSCF (Interrogating CSCF) — routes SIP to the correct S-CSCF
• S-CSCF (Serving CSCF) — core call control logic
• PCRF/PCF — policy and charging, triggers QCI 1 bearer setup

The SIP signaling path is: UE -> P-CSCF -> I-CSCF -> S-CSCF -> (peer network or same S-CSCF for on-net calls).

Any failure in this SIP chain — registration timeout, authentication error, SDP negotiation failure, proxy unreachable — prevents call establishment or causes mid-call failures.

Layer 3: Media (codec + RTP)

The actual voice audio is:

1. Encoded by a codec (AMR-NB, AMR-WB, or EVS)
2. Packetized into RTP (Real-time Transport Protocol) packets
3. Transmitted over the QCI 1 bearer
4. Received and decoded at the far end

The codec selection is negotiated in the SIP SDP (Session Description Protocol) offer/answer exchange. Common VoLTE codecs:

Codec	Bandwidth	Sample Rate	Typical Bitrate	MOS Range
AMR-NB	Narrowband (300-3400 Hz)	8 kHz	4.75-12.2 kbps	3.0-3.8
AMR-WB	Wideband (50-7000 Hz)	16 kHz	6.6-23.85 kbps	3.8-4.5
EVS	Super-wideband / fullband	Up to 48 kHz	5.9-128 kbps	4.0-4.8

AMR-WB (also marketed as “HD Voice”) is the most common VoLTE codec. EVS provides superior quality but requires both endpoints to support it.

The 10-minute diagnostic method

The diagnostic method follows a strict layer-by-layer sequence. Each step either confirms the layer is functioning correctly (proceed to next) or identifies the failure (stop and remediate).

Minute 0-2: Check IMS registration status

What to verify: Is the UE registered with the IMS network?

IMS registration is the prerequisite for all VoLTE functionality. Without a valid IMS registration, the UE cannot place or receive VoLTE calls and will fall back to CSFB (Circuit-Switched Fallback) or fail entirely.

Diagnostic steps:

1. Check NAS signaling — look for PDN Connectivity Request to the IMS APN. The UE must have an active PDN connection to the IMS APN (typically named “ims” or “ims.operator.com”).

2. Verify SIP REGISTER — the UE sends a SIP REGISTER message to the P-CSCF. Look for:

   • REGISTER sent with valid credentials (IMPI/IMPU)
   • 401 Unauthorized (challenge) received from P-CSCF
   • REGISTER resent with authentication response
   • 200 OK received — registration successful

3. Check registration expiry — the 200 OK contains an Expires header (typically 3600 seconds). If the UE fails to re-register before expiry, IMS registration lapses.

Common IMS registration failures:

Failure	SIP Response	Root Cause
Authentication failure	403 Forbidden	ISIM/USIM credentials mismatch
P-CSCF unreachable	No response to REGISTER	IMS APN routing, DNS failure
Registration timeout	408 Request Timeout	Network congestion, P-CSCF overloaded
Not authorized for VoLTE	503 Service Unavailable	HSS/UDM subscription missing VoLTE entitlement
IPSec failure	No response after 401	Security association establishment failed

HiCellTek diagnostic: HiCellTek decodes SIP messages from the Qualcomm diagnostic interface, showing the complete REGISTER exchange with timestamps and response codes. An IMS registration failure is immediately visible in the Layer 3 message view.

Minute 2-4: Verify QCI 1 bearer setup

What to verify: When a call is initiated, does the network establish a dedicated QCI 1 bearer?

The QCI 1 bearer setup is triggered by the PCRF/PCF after the SIP INVITE is processed. The sequence is:

1. UE sends SIP INVITE to P-CSCF
2. P-CSCF forwards to S-CSCF
3. S-CSCF interacts with PCRF/PCF for policy
4. PCRF/PCF triggers dedicated bearer activation
5. MME/SMF sends Activate Dedicated EPS Bearer Context Request to UE (in NAS)
6. Network sends RRCReconfiguration with DRB configuration for QCI 1
7. UE responds with Activate Dedicated EPS Bearer Context Accept

What to look for in Layer 3:

• NAS: Activate Dedicated EPS Bearer Context Request with EPS QoS showing QCI=1
• RRC: RRCReconfiguration adding a new DRB (Data Radio Bearer) with PDCP configuration for voice
• NAS: Activate Dedicated EPS Bearer Context Accept from UE

Common QCI 1 bearer failures:

Failure	Symptom	Root Cause
Bearer not established	No dedicated bearer request in NAS	PCRF policy misconfiguration, missing VoLTE service profile
Bearer rejected by UE	Activate Dedicated EPS Bearer Context Reject	UE protocol error, PDN connection issue
Bearer released prematurely	Deactivate EPS Bearer Context Request during call	Resource preemption, radio link degradation
Wrong QCI	Bearer established with QCI != 1	PCRF template error

Impact of bearer failure: Without a QCI 1 bearer, voice packets are carried on the default bearer (QCI 5 or QCI 9), which provides no GBR guarantee, higher delay budget, and no scheduling priority. The result is inconsistent voice quality — sometimes acceptable, sometimes severely degraded depending on cell load.

Minute 4-6: Analyze codec negotiation

What to verify: What codec did the endpoints agree on, and is it optimal?

Codec negotiation happens in the SIP SDP offer/answer exchange:

1. SDP Offer (in SIP INVITE): The calling UE proposes a list of supported codecs with parameters
2. SDP Answer (in SIP 200 OK or 183 Session Progress): The called party (or IMS) selects the codec

Key SDP parameters to check:

m=audio 1234 RTP/AVP 99 100 101
a=rtpmap:99 AMR-WB/16000/1
a=fmtp:99 mode-change-capability=2; max-red=0
a=rtpmap:100 AMR/8000/1
a=rtpmap:101 telephone-event/8000

Diagnostic questions:

• Does the SDP offer include AMR-WB or EVS? If only AMR-NB is offered, the UE or IMS configuration is limiting quality.
• Does the SDP answer select AMR-WB? If the offer includes AMR-WB but the answer selects AMR-NB, the far end or an intermediate IMS node is downgrading.
• What codec mode set is negotiated? AMR-WB mode 23.85 kbps provides the best quality. If mode-set restricts to lower bitrates, quality is intentionally limited.
• Is EVS offered and accepted? EVS requires both UEs and the network to support it. A codec mismatch silently falls back to AMR-WB or AMR-NB.

Common codec issues:

Issue	SDP Evidence	Impact
AMR-NB instead of AMR-WB	SDP answer selects AMR/8000	Maximum MOS limited to ~3.8
Low bitrate mode	mode-set=0,1,2 (AMR-WB)	Quality degraded vs full mode set
No codec agreement	SIP 488 Not Acceptable Here	Call fails completely
EVS to AMR-WB fallback	SDP answer has AMR-WB despite EVS offer	Quality reduced from potential maximum
DTMF handling issue	Missing telephone-event in SDP	DTMF tones not transmitted (IVR failures)

Minute 6-8: Assess RTP media quality

What to verify: Is the voice media stream being delivered with acceptable quality?

Even with correct IMS registration, a proper QCI 1 bearer, and the right codec, the actual voice quality depends on RTP packet delivery.

RTP quality metrics:

Metric	Acceptable	Marginal	Poor
Packet loss	< 1%	1-3%	> 3%
Jitter	< 20 ms	20-50 ms	> 50 ms
One-way delay	< 150 ms	150-300 ms	> 300 ms
RTCP round-trip	< 200 ms	200-400 ms	> 400 ms

Radio conditions affecting RTP:

The QCI 1 bearer’s RTP performance is directly tied to the radio layer:

• High BLER (> 2%) causes RLC retransmissions, increasing jitter and potentially exceeding the RTP discard timer
• Low SINR (< 5 dB) forces the scheduler to use robust MCS (QPSK), reducing capacity and increasing scheduling delay
• Handover interruption — during handover, there is a brief interruption (20-100 ms) where RTP packets may be lost or delayed. Multiple handovers during a call compound this effect.
• Cell edge conditions (RSRP < -110 dBm) produce unreliable packet delivery even with HARQ

Diagnostic correlation:

The key diagnostic technique is to correlate RTP quality metrics with radio KPIs on the same timeline:

[T+12.0s] SINR drops from 15 dB to 3 dB (entering building shadow)
[T+12.5s] MCS drops from 20 to 4 (scheduler adapts)
[T+13.0s] BLER increases to 8% (retransmissions begin)
[T+13.5s] RTP jitter spikes to 60 ms
[T+14.0s] MOS drops from 4.1 to 2.8
[T+16.0s] Handover to PCI 215 (A3 event triggered)
[T+16.1s] 80 ms RTP gap during handover
[T+16.5s] SINR recovers to 12 dB on new cell
[T+17.0s] MOS recovers to 3.9

This correlation is impossible to achieve with separate tools — one for radio measurement and another for voice quality. HiCellTek captures both simultaneously from the same device.

Minute 8-10: Measure MOS (Mean Opinion Score)

What to verify: What is the actual perceived voice quality as measured by an objective MOS algorithm?

MOS (Mean Opinion Score) is the definitive quality metric. While RTP jitter and packet loss are useful diagnostics, they do not directly predict what the subscriber hears. MOS integrates all impairments into a single 1-5 score.

MOS measurement methodology:

1. Reference audio: A standardized speech sample (ITU-T P.501) is played from one endpoint
2. Received audio: The audio as received at the measurement device is captured
3. Objective comparison: The algorithm compares the received audio against the reference to compute a MOS score

MOS scoring algorithms:

• ViSQOL (Google Research) — open-source, excellent performance on AMR-WB and EVS, supports speech (16 kHz) and audio (48 kHz) modes. HiCellTek integrates ViSQOL natively via JNI (libvisqol_jni.so).
• POLQA (ITU-T P.863) — industry standard for regulatory reporting, requires commercial license.

MOS interpretation for VoLTE:

MOS Score	Quality	Typical Cause if Below Target
4.0-4.5	Excellent	AMR-WB or EVS in good radio conditions
3.5-4.0	Good	Minor impairments, acceptable for most users
3.0-3.5	Fair	Noticeable degradation, codec mode adaptation active
2.5-3.0	Poor	Significant jitter, packet loss, or codec downgrade
< 2.5	Unacceptable	Severe radio degradation or IMS media path issue

Operator SLA targets: Most operators target MOS > 3.5 at 95th percentile for VoLTE. Premium operators target MOS > 3.8.

Decision tree: isolating the failure layer

After completing the 10-minute diagnostic, use this decision tree to identify the root cause:

START
  |
  v
Is IMS registration successful?
  |-- NO --> IMS layer issue
  |          Check: IMS APN, P-CSCF reachability,
  |          ISIM credentials, VoLTE subscription
  |
  |-- YES
  |
  v
Is QCI 1 bearer established?
  |-- NO --> Bearer/policy issue
  |          Check: PCRF policy, VoLTE service profile,
  |          dedicated bearer template, MME configuration
  |
  |-- YES
  |
  v
Is codec negotiation correct?
  |-- NO --> SDP/codec issue
  |          Check: UE codec support, IMS codec policy,
  |          SDP filter rules, transcoding configuration
  |
  |-- YES
  |
  v
Is RTP quality acceptable?
  |-- NO --> Radio/transport issue
  |          Check: RSRP, SINR, BLER, handover frequency,
  |          QCI 1 scheduling, uplink power headroom
  |
  |-- YES
  |
  v
Is MOS acceptable?
  |-- NO --> Codec mode or end-to-end issue
  |          Check: codec bitrate adaptation, AMR mode set,
  |          transcoding in IMS path, far-end audio quality
  |
  |-- YES --> VoLTE operating normally

Advanced VoLTE failure scenarios

Scenario 1: One-way audio

Symptom: Caller can hear the called party, but the called party cannot hear the caller (or vice versa).

Diagnostic: Check the SDP negotiation — the c= (connection) line and port numbers must be correct in both offer and answer. One-way audio typically indicates a NAT traversal failure or firewall blocking RTP in one direction. Also verify that the QCI 1 bearer is bidirectional (check UL and DL TFT — Traffic Flow Template — in the bearer activation).

Scenario 2: Call drops after handover

Symptom: VoLTE call drops consistently when the UE moves between cells.

Diagnostic: Capture the handover sequence in Layer 3. Look for:

• RRCReconfiguration with mobilityControlInfo (handover command)
• Is the QCI 1 bearer included in the handover configuration?
• Is the target cell supporting VoLTE (check SIB1 for IMS emergency indicator)?
• Is there SRVCC (Single Radio Voice Call Continuity) configured for inter-RAT handover to 3G/2G?

Scenario 3: Delayed call setup

Symptom: Long delay (> 5 seconds) between dialing and hearing ringback tone.

Diagnostic: Time each step of the SIP signaling sequence:

• INVITE sent to 100 Trying: should be < 500 ms (P-CSCF processing)
• 100 Trying to 183 Session Progress: should be < 2 seconds (end-to-end SIP routing)
• 183 Session Progress to 180 Ringing: should be < 1 second (bearer setup)

Identify which segment introduces the excessive delay. Common causes: DNS resolution delays for P-CSCF, overloaded S-CSCF, inter-operator SIP routing issues.

Scenario 4: Intermittent voice quality degradation

Symptom: Voice quality varies during the call — sometimes clear, sometimes choppy, in a repeating pattern.

Diagnostic: Correlate MOS score with radio KPIs over the call duration. Look for:

• Periodic SINR dips — indicates UE is moving through an interference pattern
• AMR mode switching — the codec adapts to radio conditions by changing bitrate. Frequent mode switches between high (23.85 kbps) and low (6.6 kbps) indicate unstable radio conditions
• Scheduling priority issues — if the cell is heavily loaded, QCI 1 scheduling may be impacted despite the GBR guarantee

HiCellTek for VoLTE diagnostics

HiCellTek integrates all three VoLTE diagnostic layers into a single real-time measurement workflow:

• IMS signaling: Complete SIP message decoding from the Qualcomm diagnostic interface — REGISTER, INVITE, BYE, SDP offer/answer
• Bearer monitoring: QCI 1 dedicated bearer setup detection, TFT analysis, bearer modification and release events
• Radio correlation: Real-time RSRP, SINR, BLER, MCS captured simultaneously with voice quality metrics
• MOS scoring: Integrated ViSQOL MOS computation with per-second scoring during active calls — see the VoLTE testing app for the full feature set
• Unified timeline: All events — SIP, NAS, RRC, L1 KPIs, MOS — displayed on a single correlated timeline
• Structured export: Complete VoLTE diagnostic data in PCAP, CSV, HLOG, or QMDL format via the integrated Layer 3 decoder

For a deeper understanding of QoE measurement methodology including video MOS, see our network QoE measurement guide. For 5G NR radio optimization techniques that directly impact VoNR quality, refer to the 5G NR optimization RF engineer guide.

Conclusion

VoLTE troubleshooting requires a systematic, layer-by-layer approach. The 10-minute diagnostic method presented here — check IMS registration, verify QCI 1 bearer, analyze codec negotiation, assess RTP quality, measure MOS — eliminates guesswork and isolates failures to the responsible layer. Each step produces actionable evidence that can be directly included in vendor escalation tickets or internal engineering reports.

HiCellTek transforms VoLTE diagnostics from a multi-tool, multi-person exercise into a single-device workflow. Real-time Layer 3 decoding, radio KPI capture, and integrated MOS measurement on a standard Android smartphone give field engineers the complete diagnostic picture in 10 minutes.

Ready to diagnose VoLTE issues in the field? Contact us at sales@hicelltek.com or visit hicelltek.com to start your evaluation.