For intermittent CPRI link faults, capture the alarm and optical readings before disturbing the connection, then check connector cleanliness, locking, cable routing and outdoor seals. Use a known-good, correctly specified jumper to isolate the suspect section. Follow an explicit rate-mismatch or radio-power alarm directly into its corresponding equipment check.
CPRI means Common Public Radio Interface. In a fiber-based installation, it connects a baseband unit (BBU) with a remote radio unit or head (RRU/RRH). This guide focuses on the passive fiber path in fiber-to-the-antenna (FTTA) networks and the evidence maintenance teams need before ordering replacement assemblies.
Before you start
Three things apply to every check in this guide, so they are stated once rather than repeated in each section:
- Work to the site's and the equipment vendor's approved procedures. Arrange anything that affects service with the network operations team first.
- Follow the site's optical safety procedure. Use an appropriate inspection probe; never look into a fiber or optical port.
- Treat every reading as evidence. Record it with a timestamp and the link identity before you change anything.
The passive route a CPRI link depends on, from the BBU patching position through the feeder, splice closure and outdoor jumper to the radio. Each section carries its own cable and port identity, and its own evidence.
1. CPRI link fault triage: start with the alarm and the affected direction
A CPRI interface alarm identifies a link problem that still needs localization. Record the exact alarm text, affected BBU and radio ports, direction and time, then confirm that the radio has power and has completed its startup sequence.
The VIAVI/JDSU CPRI fronthaul white paper describes signal, frame and remote alarm indications as useful diagnostic information. Use the equipment vendor's interpretation of those indications to select the next check:
| Initial evidence | First investigation | Record before intervention |
|---|---|---|
| Loss of signal or a low-receive-power alarm | Optical source, receive direction and intervening fiber path | Both ends' optical readings and module identities |
| Frame/synchronization errors with light present | Optical quality, supported line rate, clocking and equipment status | Error counters, synchronization status and recent changes |
| Explicit rate or port-mapping mismatch | Compare the approved configuration with installed equipment and cabling | Configured rate, module support and physical port map |
| Radio power or restart event | Radio supply and startup status | Power alarms, restart reason and correlated CPRI events |
Carry that first reading into a three-step order rather than working through the sections below in sequence. The branch you take is decided by whether light is arriving at all.
Step 1 records the alarm type and DDM history without disturbing the connection; step 2 follows the branch that fits the evidence; step 3 escalates with the measurement and change record once the passive path has been characterised.
- Read the alarm type and the digital diagnostic monitoring (DDM) history. Do not disturb the connection first. The recorded history distinguishes "no light" from "light present but errors".
- Follow the branch that fits. Signal loss or low receive power points to the optical path - work through end face, routing and seal. Light present with frame or synchronization errors points to optical quality, rate, clocking and configuration.
- Escalate when both are exhausted. If the passive path has been characterised and the errors persist, hand the measurement and change record to the equipment vendor.
Identify which interface is actually in service. The CPRI Cooperation specification index lists CPRI and eCPRI separately; an eCPRI deployment requires its own transport and synchronization workflow.
Draw the passive route from the BBU patching position through any feeder, splice enclosure and outdoor jumper to the radio, and record each section with the cable and port identities. A numbered, photographed route sketch is what lets a second technician reach the same conclusion from the same evidence. Glory Optical's 5G FTTA fiber connection solution shows how the cable, closure and connector positions in that route are typically grouped.
2. What makes a CPRI link fail intermittently
A permanent break is straightforward to find. An intermittent fault is harder because the link is functional whenever it is measured. Most of these faults share a pattern: a condition that sits close to a limit, combined with something that changes - temperature, load, wind, vibration or moisture.
These are the categories worth checking on the passive path and its immediate interfaces. They are not listed in order of likelihood, and more than one can be present at the same time:
| Category | Mechanism | What the evidence tends to look like |
|---|---|---|
| Marginal optical budget | Receive power sits close to the receiver's limit, so a small additional loss or drift crosses it | Receive power inside the specified range but with little margin; faults that correlate with temperature or daytime load |
| End-face contamination | A particle over the core changes loss as the joint settles or moves | Fault appears after a mating event, cabinet access or work nearby; link passes again after cleaning |
| Incomplete engagement | The latch, bayonet or threaded coupling is not fully seated, so vibration or thermal movement opens the joint | Fault follows wind, traffic load or handling; the link recovers when the connector is touched or reseated |
| Strain or bend violation at a dynamic location | A door, tray, clamp or entry edge loads the fiber and the load changes with temperature or wind | Loss changes when the cable is moved; flattening, tight ties or abrasion visible on inspection |
| Moisture ingress | A seal, gasket or cable entry has failed, and water in the assembly changes loss until it dries | Fault correlates with rain or humidity and clears in dry conditions; deposits or staining found on opening |
| Rate or module mismatch | The installed module does not support the configured CPRI option rate, or supports it only marginally | The link comes up and then drops, or error counters rise; the alarm text names rate or mapping rather than signal loss |
| Module seating or thermal behaviour | The module is not fully latched in the cage, contacts are oxidised, or the module reaches its operating limit inside a hot cabinet | Fault correlates with cabinet temperature; DDM temperature reading is high; reseating helps temporarily |
| Reflection or return-loss condition | A damaged, mismatched or non-mating end face produces reflections that affect the received signal quality | Frame or synchronization errors rather than a clean loss of signal |
Two consequences follow from this list. First, a correlation is not a diagnosis: a fault that follows rain is evidence about when to inspect, not proof that moisture is the cause. Second, an intermittent fault that disappears when a component is handled has produced useful evidence, not a repair - the handling may simply have changed the joint at the same time.
For the fiber-side design context behind these categories, see Glory Optical's overview of building a reliable FTTA fiber foundation.
3. Capture CPRI optical readings while the fault is present
Compare receive power with the operating limits of the exact installed optical module, at its supported rate and wavelength. Collect readings from both ends during healthy and faulty periods so the team can assess changes against the same link state.
Small form-factor pluggable modules (SFPs) may expose digital diagnostic monitoring, or DDM, including transmit/receive power and temperature; the monitoring registers are defined by SFF-8472. Record the units and collection interval. A brief healthy reading between outages can miss the event being investigated.
Use a shared record rather than separate screenshots without timestamps:
| Field | What to capture |
|---|---|
| Link identity | Site, sector, BBU slot/port, radio port and fiber IDs |
| Module identity | Part number, serial number, supported rate and wavelength at each end |
| Optical conditions | Tx/Rx readings, applicable receiver limits, temperature and any module alarms |
| Fault timing | Alarm onset/clear times, counter changes, radio restart and configuration changes |
| Physical context | Rain, temperature change, cabinet access or recent cable work, if observed |
| Intervention | Exact component touched or replaced, time, and before/after result |
A receive-power change helps identify a direction to investigate; check its associated transmitter, module and fiber path together. Stable sampled power still leaves questions about brief transients, reflections, module behavior and frame synchronization - which is why an unremarkable DDM trace narrows the search rather than closing the optical path as a cause.
Illustrative receiver-margin check
Suppose an approved receiver's specified range is −14 to −3 dBm and the measured receive power is −13.2 dBm. The reading is 0.8 dB above the lower limit. That small measured separation merits investigation against the project's required margin and measurement uncertainty, even though the reading lies within the stated range.
Projects commonly plan a margin above the receiver's minimum sensitivity - a figure in the region of 3 dB is often quoted - but the value used for acceptance is the one written in the project's own link budget. Compare the measured reading against three references in turn: the module's specified range, the calculated link budget, and the planning margin. A reading that passes the first test can still fail the third.
These are example inputs, not CPRI-wide limits or Glory Optical specifications. Apply the actual module data and link budget. The EXFO mobile fronthaul testing guide provides the broader optical-budget and CPRI test context.
4. Inspect FTTA connector end faces and locking parts
Inspect both sides of each suspect optical connection, clean contamination with the appropriate tools, and inspect again before mating. Check the mechanical latch or coupling separately so the service record distinguishes end-face condition from connection retention.
IEC 61300-3-35:2022 addresses visual inspection of connector end faces and fiber-stub transceivers. Its scope explicitly separates visual inspection from measurements such as attenuation and return loss, and its acceptance criteria are organised by zones on the end face rather than as a single pass/fail figure. Because those criteria depend on the connector type, the fiber and the required grade, record which criteria apply to the specific interface on site. Record both kinds of evidence where the acceptance plan requires them.
Use this sequence for an accessible suspect connection:
- Record the port and jumper identity, then disconnect under the approved procedure.
- Inspect both mating end faces using the correct probe tips. Clean any contamination and reinspect.
- Check the latch, adapter alignment and protective housing for damage or incomplete engagement.
- Reconnect using the manufacturer's method, support the cable, and record the optical and alarm response.
VIAVI's Inspect Before You Connect guidance emphasizes checking both sides. For outdoor assemblies, also inspect the coupling and seal without transferring dirt onto a cleaned ferrule. Glory Optical's connector basics guide covers how ferrule, alignment and polish contribute to a mating interface.
5. Trace the fiber jumper through bends, clamps and entry points
Trace the jumper through its installed route and compare bends, pulling exposure and support with the cable manufacturer's limits. Pay particular attention to locations where a door, tray, clamp or cable entry can change the load on the fiber.
Look for flattened jacket sections, tight ties, abrasion, sharp entry edges and unsupported cable weight near the connector. Check slack storage with the tray and enclosure in their normal closed positions. Photograph the condition before correcting it, and replace visibly damaged material according to the site's repair criteria.
The FOA cable installation guide puts cable-specific bending and pulling limits at the center of handling practice. Use those limits for the complete outdoor assembly, including its connector boot and strain relief.
VIAVI's wireless fiber deployment guidance describes how contamination and bending can produce marginal as well as permanent faults. When an outage follows cabinet access or weather, use that timing to choose inspection points and collect measurements. Preserve it as an observed correlation until testing identifies the faulty section.
For detailed routing, cleaning and testing procedures, see Glory Optical's fiber patch cable installation guide.
6. Check outdoor jumper seals as a complete assembly
Inspect the enclosure, connector housing, cable entry and retention hardware together. A repair should restore the specified sealing configuration and remove the mechanical condition that allowed movement, contamination or exposure.
Open outdoor connection points under suitable working conditions and the approved maintenance procedure. Record moisture, deposits, damaged seals and cable movement before cleaning or replacing parts. These observations guide the investigation; the optical tests establish their effect on the link.
| Location | Inspect | Required repair detail |
|---|---|---|
| Weatherized connector | Coupling engagement, seal condition and mating interface | Exact housing, receptacle and seal configuration |
| Cable gland or gel entry | Cable diameter/profile, seating and unused-port plugs | Approved cable-to-entry combination |
| Enclosure perimeter | Gasket path, trapped material, damage and latch closure | Replacement parts and closure procedure |
| Internal routing | Fiber clearance, supported cable and stored slack | Routing that remains correct when closed |
IEC 60529 classifies enclosure ingress protection. For the replacement assembly, request evidence identifying the tested configuration, including mated or capped state and cable entries. Specify temperature, sunlight exposure and mechanical retention alongside ingress requirements.
A single closure commonly carries entries with materially different rated cable ranges. CommScope's HFDC hardened fiber distribution closure documentation, for example, distinguishes trunk from jumper entries: the Installation Bulletin 7860418 revision in circulation lists four trunk ports up to 17.5 mm and twelve jumper ports up to 5.5 mm. That is roughly a threefold difference in rated diameter on the same product. It is the clearest illustration of why the cable-to-entry combination has to be confirmed per port rather than per product - CommScope's fiber closure documentation is where the current revision for a specific product should be confirmed. Its dimensions and tightening instructions apply to that product; use the corresponding instructions for the enclosure on site.
Where the route contains a protected cable-to-cable splice point, Glory Optical's GL-H03 fiber optic junction box is a catalog option to evaluate. It is a splice enclosure, so confirm cable-entry fit, internal storage and mounting for that location rather than treating it as a radio-port connector housing. For the wider FTTA cable, closure and connector set, see Glory Optical's FTTA cable, closure and connector guide, and for the field-installed option at an outdoor port, the IP68 slim 3-in-1 outdoor connector guide.
7. Isolate the jumper, SFP and equipment port
Change one controlled variable at a time and record what happens under comparable observation conditions. Start with an accessible suspect jumper when the evidence points to that section, using a known-good assembly that matches the required optical and mechanical interfaces.
| Controlled change | What the result suggests | What still needs confirmation |
|---|---|---|
| Replace only the suspect jumper | Improvement focuses attention on the removed jumper and its disturbed interfaces | Reseating or cleaning may have changed the connection at the same time |
| Substitute an approved equivalent SFP | A fault that follows the module supports a module-related diagnosis | Module compatibility, seating and repeatability |
| Test a suspect fiber section independently | Excess loss or a localized event focuses the physical repair | Test references, wavelength and event location |
| Move to an approved alternative equipment port | A repeatable change can help distinguish port from path | Correct configuration, mapping and vendor procedure |
Label and retain suspect components for evaluation. A temporary recovery after handling is useful evidence, but an intermittent fault may need a longer observation period before the repair can be accepted.
The configured CPRI rate and the module that supports it
Check the configured CPRI rate, supported optics and physical port mapping whenever the alarm or recent changes point there. A generic "10G" label is insufficient to establish support for the required CPRI option, because the CPRI line bit rates are their own set of values. They are not the same as the 10.3125 Gbps of 10 Gigabit Ethernet:
| CPRI option | Line bit rate |
|---|---|
| Option 1 | 614.4 Mbps |
| Option 2 | 1228.8 Mbps |
| Option 3 | 2457.6 Mbps |
| Option 4 | 3072.0 Mbps |
| Option 5 | 4915.2 Mbps |
| Option 6 | 6144.0 Mbps |
| Option 7 | 9830.4 Mbps |
| Option 8 | 10137.6 Mbps |
These values are the CPRI line bit rates defined in the published CPRI specification; confirm the option and revision that apply to the equipment on site.
Modules sold for this application are therefore specified as multi-rate, typically quoted as a range that covers several options rather than a single Ethernet data rate. Three questions settle whether an installed module is right for the link:
- Does the module's published supported-rate range include the exact CPRI option configured on the port?
- Does the module's rated wavelength and reach match the fiber and distance installed?
- Is the module on the equipment vendor's approved list for that port and software release?
Ask for the module datasheet rather than relying on the speed printed on the label. Avoid trial configuration changes on a live radio without the vendor's procedure. Glory Optical's guide to FTTA connector compatibility and RRU selection covers how the radio-side interface is matched.
When the passive path has been characterized and errors persist, escalate with the measurements and change log. Ask the equipment vendor to assess module/port behavior, synchronization, approved firmware and relevant known issues. AMD's CPRI-specific checks illustrate the importance of transceiver settings and clock status at the implementation level; their register instructions are for that IP core, not a universal base-station repair procedure.
8. Specify the replacement FTTA assembly
Order the replacement against both equipment interfaces and the complete cable construction. Give procurement enough information to compare the proposed assembly with the approved installation, including its sealing and retention parts.
| Replacement field | Include in the request |
|---|---|
| Equipment interfaces | Radio/BBU model, port drawing, connector format, polish and any keyed outer housing |
| Optical path | Fiber type/count, Tx/Rx mapping, length and project optical limits |
| Mechanical construction | Jacket, cable diameter, armor or strength members, bend and load requirements |
| Environment | Installation temperature, operating exposure and required sealing configuration |
| Accessories | Glands, seals, caps, clamps, breakout lengths and mounting provisions |
| Evidence | Drawing revision, sample-fit approval, optical report fields and qualification documents |
| Supply control | Quantity by configuration, spare quantities, labeling, lead time and substitution approval |
Match the housing as well as the optical connector
Glory Optical lists an ODVA LC duplex outdoor patch cord and a FullAXS LC fiber patch cable. These are separate catalog options to review against the equipment interface. Confirm the mating housing, keying, engagement and cable-entry configuration even when both descriptions include LC. Two assemblies can both be described as LC and still not mate, and the environment field above is where the rated temperature range of the proposed assembly should be confirmed in writing - not inferred from a general "outdoor" description. The difference between an assembly built for a protected indoor cabinet and one built for a tower-mounted port is set out in Glory Optical's comparison of outdoor and indoor fiber patch cords.
For a project-specific replacement, use Glory Optical's OEM/ODM configuration service to discuss the drawing, length, marking and packing. Request written confirmation of the offered configuration and its supporting documents before bulk purchase.
9. Close the CPRI repair with optical and service evidence
Accept the repair against the project's optical limits and a documented period of stable equipment operation. Keep before-and-after records so the next technician can see what changed and which failure conditions were covered.
Measure the specified link's insertion loss using the agreed wavelengths, test references and compatible cords. The FOA reference on testing installed cable plant explains why the reference method matters, and covers the procedures for both singlemode (OFSTP-7) and multimode (OFSTP-14) installations - an FTTA link is a singlemode application, so the singlemode procedure and its reference cables apply. Record the boundary being measured, especially when the repair affects only one jumper within a longer path.
Use an optical time-domain reflectometer (OTDR) when event localization is needed. Short FTTA sections require suitable resolution and launch/receive arrangements, as explained in the FOA OTDR guide. Save instrument settings with the traces.
A passing OTDR trace is not on its own an acceptance test. It is a measurement taken at one time, with pulse settings chosen for the link, and it cannot show a condition that depends on movement, temperature or moisture being present. Keep it alongside the end-face results rather than in place of them. For the wider maintenance sequence around deployment and testing, see Glory Optical's deployment, testing and maintenance guide.
The handover should include:
- Cable, module and port identities, including any replacements.
- End-face inspection results, routing/sealing photographs and the approved drawing revision.
- Optical measurements, applicable limits and test setup.
- Alarm history, available error-counter trends and radio/service status after repair.
- Observation duration, conditions encountered and any remaining follow-up action.
Choose the observation window with network operations around the original recurrence pattern. There is no universal figure that applies to every intermittent fault: a fault that recurs with a weather or temperature cycle needs observation through comparable conditions, while a fault with no identified pattern needs a window long enough to cover the original interval between events. A short successful reconnect only establishes immediate recovery.
Frequently asked questions about intermittent CPRI link faults
These questions cover the four situations that most often stall an FTTA fault investigation: normal-looking optical power, a rate question, the observation window, and a link that recovers as soon as it is touched.
The CPRI link is intermittent but the optical power looks normal. Where should I look first?
Read the DDM history rather than the current reading. If no power excursion lines up with the fault times, the evidence points away from the passive path and towards rate, clocking, module or port behaviour. Treat that as a direction, not a conclusion: sampled DDM can miss short transients, so an unremarkable trace narrows the search without clearing the optical path.
Can a standard 10G SFP+ run a CPRI Option 7 or Option 8 link?
Not on the strength of the "10G" description alone. 10 Gigabit Ethernet runs at 10.3125 Gbps, while CPRI Option 7 is 9830.4 Mbps and Option 8 is 10137.6 Mbps. The module has to be specified for the exact CPRI option rate configured on the port, and it should appear on the equipment vendor's approved list. Ask for the datasheet.
How long should an intermittent fault be observed before the repair is accepted?
There is no single figure that fits every case. Set the window from the observed recurrence pattern: a weather-associated fault needs comparable conditions, and a fault with no clear pattern needs a window at least as long as the shortest interval previously observed between events. Agree the window with network operations before the repair is signed off.
Does a passing OTDR trace rule out the fiber path?
No. An OTDR measures at one moment, at a chosen wavelength and pulse width. A marginal, movement-dependent or moisture-dependent condition may be absent while the test is running. Use OTDR results to localize events, and keep them alongside end-face inspection and the fault-time optical records.
The link came back as soon as I reseated the connector. Is the repair finished?
Record it as evidence, including the before-and-after reading and the exact component touched, but do not treat it as an accepted repair. Reseating changes the joint at the same time as it may remove the cause, and an intermittent fault can return days later under the conditions that produced it.
Preparing a replacement order? Send Glory Optical your FTTA replacement requirements with the interface drawings, cable route, failure observations and quantities. Request a proposed passive assembly, matching accessories and the test-document package needed for your maintenance approval.
GLORY OPTICAL · PRODUCT SELECTION
Products to evaluate along the CPRI passive path
These catalog options cover the four positions a CPRI link passes through, from the BBU patching position to the radio port. Confirm the mating interface, port and sealing configuration for the exact assembly before approval.
OUTDOOR JUMPER
Outdoor Single Mode Fiber Patch Cord
G.657.A2 OS2 single mode in a 5.0 or 7.0 mm duplex construction, with aramid strength members and a steel armor option. The published table lists IP67/IP68 at the connector, a −40 °C to +75 °C operating range and an armored tensile rating of ≥1000 N.
Specify: Confirm the connector format and polish at each end, the jacket and armor option, the installed length and the bend limits that apply to the tower-side route.
View product specifications →RADIO-SIDE JUMPER
FullAXS LC Fiber Patch Cable
A 9/125 µm single mode duplex assembly with LC duplex connectors. The published table gives a 4.8 mm jacket, a 20 mm bending radius, 450 N installation tensile strength, 3000 N/100 mm crush resistance and a maximum insertion loss of 0.30 dB.
Specify: Confirm that the hardened housing, keying and engagement match the radio port, and record the rated temperature range for the exposed mounting position.
View product specifications →SPLICE ENCLOSURE
Fiber Optic Splice Closure
An in-line closure for the protected splice point on the route. The published table lists PC-PBT material, IP68 mechanical sealing, up to 144 fiber splice capacity across three trays, and 20 cable ports with mixed diameters for aerial, wall, manhole or direct-burial mounting.
Specify: Match each feeder and jumper cable to a specific port diameter, confirm tray and slack capacity, and confirm the mounting method for the location.
View product specifications →BBU PATCHING
ODF Fiber Optic Patch Panel
A rack-mounted optical distribution frame for the BBU-side patching position, where the port map in section 7 is recorded. The published GL-JPF series table lists 1U/2U sizes, SC, FC, ST or SC duplex adapter options, and cable entry diameters from Φ8–16 mm (1U) to Φ8–25 mm (2U).
Specify: Confirm the adapter type and port count against the recorded fiber map, and confirm the cable entry diameter and rack unit height.
View product specifications →
