The Unique Aspects of Telecom Infrastructure Monitoring
Radio sites (BTS, eNodeB, 5G gNodeB) are critical infrastructure whose failure immediately affects thousands of users. The pressure to maintain availability is extreme: contractual SLAs between mobile operators and TowerCos (tower operators) require availability levels of 99.9% to 99.99% per year, which translates to 8.7 hours and 52 minutes of permitted downtime per year, respectively.
Furthermore, these sites are often isolated (rural areas, rooftops, transmission towers), difficult to access, and unmanned, making remote monitoring essential to achieving these performance levels.
Components of a Supervised Radio System
| Category | Equipment | Monitored Parameters |
|---|---|---|
| Power Supply | Rectifier, Inverter | Voltage, Current, Load (%) |
| Backup Power | VRLA/Li-Ion Batteries | Voltage, Temperature, Capacity, BMS |
| Generator | Diesel Generator | Fuel, Hours, Voltage, Start-up |
| Active Equipment | BBU, RRU, router | Temperature, CPU, internal alarms |
| Transmission | Microwave, fiber | RSL, throughput, BER, alignment |
| Environment | Cabinet cooling, hygrometer | Temperature, RH |
| Security | Door sensors, PIR, camera | Intrusion, access |
Multi-site Telecom Monitoring Architecture
The monitoring of a network of radio sites is based on two levels:
Site level: One IoT gateway per site collects all local data (power supply, batteries, generator, environmental conditions, security) and transmits it to the cloud platform via a dedicated SIM card provided by a third-party carrier.
Network Level: Active equipment (base stations, routers) reports its alarms via SNMP or proprietary protocols to the operators’ NMS systems, which can be integrated into the infrastructure monitoring platform for comprehensive correlation.
This two-tier architecture ensures that infrastructure alarms (power, environment) remain visible even when the network equipment itself is down.
Energy Incident Management
The management of a power outage at a radio site follows a specific sequence:
- T+0: EDF power outage detected → Automatic switchover to batteries → Alert sent to the energy team and the site operator.
- T+5 min: If outage > 5 min → Automatic startup of the generator → Monitored startup confirmation (or alert if startup fails).
- T+30 min: Status report sent: remaining battery life, generator status, estimated duration of outage.
- T+2 h: If outage > 2 h → Escalation alert, fuel supply check, scheduling of EDF intervention.
- Return to EDF grid: Switchover back to grid power → Generator shutdown → Incident closure report.
Preventive Battery Maintenance
Backup batteries are the Achilles’ heel of telecom sites: they age silently until the day they can no longer provide the required runtime. Monitoring makes it possible to detect this aging before a failure occurs:
- Float voltage monitoring: An abnormally low voltage indicates a faulty cell.
- Automatic discharge test: Once a month, the platform initiates a controlled discharge test and measures the actual capacity returned.
- Temperature monitoring: Every additional 10 °C halves the service life of lead-acid batteries.
- Cycle counting: Li-ion batteries have a limited number of charge/discharge cycles.
This data is used to develop a predictive replacement plan that prevents battery failures during operation.