Network assurance
Route Quality
Every destination on the VOLTA network carries a published grade, a measured threshold and an automatic remediation path. This policy describes how those grades are defined, how they are tested and what happens the moment a route falls below them.
42k
Daily test calls
< 60s
Automatic reroute
4.31
Median network MOS
99.999%
Core availability
Grades
Three grades, published thresholds
Platinum
Direct operator or tier-1 IPX routes with guaranteed CLI, no transcoding and no intermediate resellers. Reserved for enterprise, contact-centre and wholesale premium traffic.
- ASR ≥ 62%
- ACD ≥ 3.2 min
- PDD ≤ 1.6 s
- MOS ≥ 4.20
- Loss ≤ 0.05%
Gold
Single-hop carrier routes with CLI presented on more than 97% of attempts. Standard grade for retail termination and most bilateral wholesale exchanges.
- ASR ≥ 55%
- ACD ≥ 2.6 min
- PDD ≤ 2.2 s
- MOS ≥ 4.00
- Loss ≤ 0.15%
Silver
Cost-optimised routes for price-led A-Z traffic. CLI is best-effort and transcoding may occur. Never used for enterprise SIP trunks or premium contracts.
- ASR ≥ 45%
- ACD ≥ 1.8 min
- PDD ≤ 3.2 s
- MOS ≥ 3.70
- Loss ≤ 0.40%
Metrics
What we measure, and why it matters
Quality is not a single number. VOLTA grades on four correlated families of measurement so that one healthy-looking statistic cannot hide a broken route.
ASR — Answer Seizure Ratio
Measured per destination prefix over rolling 15-minute and 24-hour windows, excluding user-busy and no-answer causes so that genuine network failures are not masked by consumer behaviour.
ACD — Average Call Duration
A durable signal of false answer supervision. Any prefix whose ACD falls more than 35% below its 30-day baseline is quarantined automatically pending a manual route review.
PDD — Post Dial Delay
Timed from INVITE to first provisional ring indication at the ingress SBC. Routes exceeding their grade threshold three consecutive intervals are demoted in the LCR table.
MOS, jitter and loss
RTCP-derived MOS is sampled continuously across every media plane. Jitter above 12 ms or loss above grade threshold triggers a media-path reroute before customers observe degradation.
Methodology
How a degradation is found and cleared
01
Continuous active testing
More than 42,000 automated test calls are placed every day from probes in nine regions to live handsets and answering machines on each active destination, verifying audio, CLI delivery and DTMF pass-through.
02
Passive CDR analytics
Every production CDR feeds the routing engine within 90 seconds. Statistical models compare each prefix against its own baseline rather than a global average, so seasonal and regional patterns do not cause false alarms.
03
Automatic reroute
When a route breaches threshold, least-cost routing demotes it and traffic shifts to the next qualifying carrier in under 60 seconds. No human approval is required to protect quality.
04
Carrier remediation
The affected carrier receives an evidence pack — sample call IDs, SIP traces and RTCP statistics — within one hour. Restoration requires 24 hours of clean test results before the route re-enters production.
Commitments
What partners can hold us to
No false answer supervision
Routes that bill unanswered calls are removed permanently, not warned. Detection runs on every prefix daily and the finding is shared with the originating partner.
CLI integrity
On Platinum and Gold routes the presented A-number is passed unmodified. Where a destination regulator mandates local presentation, substitution is disclosed on the rate sheet before traffic is accepted.
Published measurements
Partners receive weekly quality reports covering ASR, ACD, PDD, MOS and NER per destination, including quarters where our own performance regressed.
Remediation SLA
Priority-one quality incidents are acknowledged in 15 minutes, with a root-cause statement inside four hours and a written post-incident review within five business days.