INCIDENT MODEL / INC-28491
Reference CPE telemetryEU-CENTRAL · REFERENCE CASE
Reference case · Fault isolation

Diagnosis with supporting and exclusion evidence

Deterministic rules combine topology, radio path and traffic context. The review exposes contributing signals, ruled-out domains and conditions that would change the result.

Review action policy
Customer symptom
“Video calls keep dropping in the bedroom.”
Experience score
54/100
High degradation
Diagnosis confidence
95%
Ruleset RCA-2.4

Diagnosis decision

Ranked from normalized signals and topology context
Evidence threshold met
Primary root cause
Weak mesh backhaul + poor device steering

The bedroom relay path scores 38/100 and the laptop stayed attached to it despite a materially stronger living-room path.

Secondary contributor
Upload bufferbloat

Upload latency reaches 186 ms during a real-time call while the camera also sends continuous uplink traffic.

Weak mesh backhaul & poor steering
confirmed
Backhaul score 38/100 · Steering quality 57/100 · Retry rate 28%
53/100
Upload bufferbloat
contributor
Upload latency 186 ms · Concurrent real-time uplink traffic
72/100
WAN line degradation
ruled-out
Gateway latency 16 ms · WAN loss 0.1%
6/100
Router overload
ruled-out
CPU 38% · Memory 52%
8/100
Firmware regression
ruled-out
Cohort abnormal failure rate 0.7% · Current release is inside control limits
5/100

Why this diagnosis is credible

Counterfactual review

Why not WAN?

ISP gateway latency is 16 ms and WAN loss is 0.1%. Both remain stable while the bedroom path degrades.

Why not router overload?

CPU is 38% and memory is 52%. No forwarding, DHCP, or radio-health fault is present.

Why not firmware?

The firmware cohort's abnormal failure rate is 0.7%, inside its control limit, with no correlated incident spike.

What evidence would change the diagnosis?

Falsification conditions

WAN fault emerges

Gateway loss exceeds 1% across wired and wireless clients, or WAN latency rises above 60 ms.

Router resource saturation

CPU or memory exceeds 85% with correlated forwarding delay and multi-path impact.

Firmware cohort signal

Same release exceeds a 4% abnormal failure rate across comparable CPE cohorts.

Client-only limitation

Other bedroom clients stay healthy and the laptop remains degraded on the strongest path.

Observed signals

6 abnormal
Bedroom RSSIRule < -67 dBm-72 dBm
Retry rateRule > 15%28%
Backhaul qualityRule < 5538/100
Upload latencyRule > 100 ms186 ms
JitterRule > 20 ms34 ms
Channel utilizationRule > 65%82%

Ruled-out causes

3 excluded
WAN line degradation
Gateway latency 16 ms · WAN loss 0.1%
Router overload
CPU 38% · Memory 52%
Firmware regression
Cohort abnormal failure rate 0.7% · Current release is inside control limits

Recommended remediation

2 selected
01Steer laptop to living-room meshguarded
02Apply temporary video-call QoSguarded
If unresolved

Escalate to guided placement or technical support with the topology and exclusion evidence attached. Commercial impact must be measured from linked care and dispatch outcomes.