Measurement, Testing, Reliability and Standards

Introduction and Context

Optical communication systems are commissioned, operated, and maintained through measurement. Design calculations may predict link performance, but only measurement can establish whether installed transmitters, fibers, connectors, receivers, and protocols meet their intended specifications. Measurement is therefore not an auxiliary activity; it is part of the engineering life cycle.

A modern optical network may include single-mode or multimode fiber, directly modulated or coherent transmitters, wavelength-division multiplexing (WDM), optical amplifiers, reconfigurable add–drop multiplexers, and high-speed receivers. Each element introduces possible impairment: attenuation, reflection, chromatic dispersion, polarization effects, nonlinear distortion, noise, crosstalk, connector contamination, and intermittent physical damage. Reliable testing must identify these impairments, distinguish equipment limitations from installation defects, and provide traceable evidence for acceptance and maintenance.

This chapter develops the principles and procedures for optical power measurement, optical spectrum analysis, optical time-domain reflectometry (OTDR), bit-error-rate (BER) testing, and eye-diagram analysis. It also addresses commissioning, calibration, fault diagnosis, reliability engineering, standards, uncertainty, safety, and contemporary automated testing.

Learning Objectives

After studying this chapter, the reader should be able to:

1. Select and correctly use optical power meters, sources, spectrum analyzers, OTDRs, BER testers, and eye-diagram instruments.

2. Calculate link loss, power margin, return loss, and BER-related quantities under stated assumptions.

3. Develop a commissioning and acceptance-test procedure for an optical link.

4. Interpret OTDR traces, optical spectra, eye diagrams, and BER measurements.

5. Apply calibration, uncertainty, maintenance, and reliability principles.

6. Relate engineering practice to relevant ITU-T, IEEE, IEC, and safety standards.

7. Diagnose faults systematically while recognizing the limitations of each instrument.

1. Measurement Architecture and Terminology

A measurement system consists of a measurand, sensor or receiver, signal-conditioning path, reference standard, data-acquisition system, and interpretation method. In optical testing, the measurand might be launched power, received power, insertion loss, optical return loss, wavelength, spectral power density, pulse broadening, or BER.

Important terms include:

• Absolute optical power: Optical power measured relative to an instrument's calibrated detector response, normally expressed in dBm or watts.
• Insertion loss: The reduction in transmitted power caused by inserting a component or link. It is normally expressed in dB.
• Return loss: The ratio, in dB, between forward power and reflected power. Larger positive return loss indicates less reflection.
• Reflectance: Reflected power divided by incident power, usually expressed as a ratio or a negative dB quantity.
• Dynamic range: The difference between the strongest and weakest signals that an instrument can measure with specified accuracy.
• Resolution bandwidth (RBW): The effective bandwidth of a spectrum analyzer's measurement filter.
• Dead zone: A region near a strong OTDR event in which subsequent events cannot be resolved accurately.
• BER: The ratio of erroneous received bits to total received bits under defined test conditions.
• Eye opening: The vertical and horizontal clearance in an eye diagram, indicating noise and timing margin.
• Traceability: An unbroken chain of calibrations relating a measurement to recognized national or international standards.

Decibel quantities require careful reference identification. For powers P1 and P2, the ratio in decibels is

L = 10 log10(P2/P1).

If P2 is lower than P1, L is negative. Optical link loss is conventionally reported as a positive quantity:

Loss = 10 log10(Pin/Pout),

where Pin and Pout are powers in the same units and measured under comparable conditions. For absolute power referenced to 1 mW,

P(dBm) = 10 log10[P(mW)].

These expressions assume power measurements are averaged consistently and that the signals are sufficiently stable for the measurement interval.

2. Optical Power Meters and Loss Measurement

2.1 Operating principle

An optical power meter uses a calibrated photodetector, commonly silicon for shorter wavelengths and InGaAs for the principal telecommunications bands near 1310 and 1550 nm. The detector converts optical energy into an electrical signal. The meter applies wavelength-dependent calibration factors and displays power in watts, dBm, or dB relative to a reference.

Power meters are used for transmitter output power, receiver input power, point-to-point insertion loss, splitter ratios, amplifier gain, and acceptance testing. Their accuracy depends on wavelength, connector type, detector linearity, polarization, optical modulation, and cleanliness.

2.2 Reference method for insertion loss

A standard loss test uses a calibrated optical source and power meter. First, the source is connected to the meter through reference test cords and a reference level is established. The device or link under test is then inserted without changing the reference arrangement. The difference between the reference and measured level is the insertion loss.

For example, if the reference is -2.00 dBm and the connected link produces -5.60 dBm, then

Loss = (-2.00) - (-5.60) = 3.60 dB.

The arithmetic is valid because both values use the same dBm reference. The test should specify whether one-cord, two-cord, or three-cord referencing is used. Different reference methods include or exclude different connector interfaces and consequently produce different acceptance results.

2.3 Link power budget

A received-power estimate can be written as

Prec = Ptx - Lfiber - Lconnectors - Lsplices - Lsplitters - Lmargin,

when all terms are expressed in dB and Ptx and Prec are in dBm. Here, Lmargin is not a physical loss but a design allowance. For a fiber span of length L, fiber attenuation is often approximated by

Lfiber = alpha L,

where alpha is attenuation in dB/km and L is in km. This model is adequate for a first-order budget, but wavelength dependence, temperature, repair activity, aging, and component tolerances must be considered in a production design.

Suppose a transmitter launches +2 dBm into a 40 km span. Fiber attenuation is 0.22 dB/km, there are two connector pairs with total loss 1.2 dB, and splices contribute 0.4 dB. The physical loss is

(0.22)(40) + 1.2 + 0.4 = 10.4 dB.

The estimated receiver input is +2 - 10.4 = -8.4 dBm. If the receiver sensitivity is -18 dBm, the unallocated power margin is 9.6 dB. Whether this is acceptable depends on the specified engineering margin, aging model, and overload limit of the receiver.

2.4 Good measurement practice

The operator should inspect and clean every connector, allow sources to stabilize, select the correct wavelength, avoid sharp fiber bends, and use launch conditions appropriate to the system. An optical power meter should not be used to infer signal quality: a correct average power can coexist with severe dispersion, noise, or bit errors. For live networks, the test wavelength and measurement direction must be selected so that service is not disrupted.

3. Optical Spectrum Analyzers

An optical spectrum analyzer (OSA) measures optical power as a function of wavelength or optical frequency. It may use a scanning monochromator, a diffraction-grating system, or an interferometric architecture. OSAs are essential for WDM channel verification, laser wavelength measurement, side-mode suppression, amplified spontaneous emission (ASE), channel power, wavelength drift, and crosstalk.

3.1 Principal specifications

The most important specifications are wavelength accuracy, wavelength resolution, amplitude accuracy, dynamic range, sensitivity, scan speed, and polarization dependence. RBW determines the ability to separate closely spaced channels. A narrow RBW improves spectral resolution but generally increases measurement time and may reduce sensitivity. The displayed noise floor must be sufficiently below the signal under examination.

For a WDM system, an OSA measurement should record channel center wavelength, channel power, occupied bandwidth, adjacent-channel leakage, and the presence of spurious lines or ASE. Comparing channel powers without accounting for RBW can be misleading: a narrow-line signal and broadband noise distribute power differently across the analyzer filter.

3.2 Interpretation

A drifting center wavelength may indicate laser temperature-control failure, aging, or an incorrect wavelength reference. Unequal channel power may result from multiplexer loss variation, amplifier gain tilt, connector loss, or transmitter malfunction. Broad ASE elevation after an optical amplifier indicates noise accumulation, while unexpected narrow lines may indicate reflections, parasitic lasing, or crosstalk.

An OSA cannot by itself establish BER compliance. Spectral conformity and temporal/data quality are related but distinct measurements. A signal can have an apparently acceptable spectrum and still fail because of timing jitter, dispersion, nonlinear distortion, or receiver problems.

4. OTDR Measurement and Trace Analysis

4.1 Principle

An OTDR launches short optical pulses into a fiber and measures returned Rayleigh backscatter and discrete Fresnel reflections as a function of time. Distance is inferred from round-trip propagation time:

z = ct/(2n),

where z is the event distance, c is the speed of light in vacuum, t is measured round-trip time, and n is the group refractive index. The factor of two occurs because the pulse travels to the event and back. The equation assumes a known effective group index and neglects small instrument and event-location corrections.

The gradual backscatter trace reveals distributed attenuation. A splice or bend generally appears as a step or slope change, while a reflective connector or open end produces a sharp peak. A non-reflective break may appear as a sudden loss without a large reflection.

4.2 OTDR parameters

Pulse width controls a fundamental trade-off. A short pulse provides better spatial resolution and smaller event dead zones but less energy and therefore less range. A long pulse improves dynamic range but can merge nearby events. Averaging reduces random noise approximately according to the measurement architecture and averaging time, but it cannot remove a systematic error or resolve events hidden inside a dead zone.

The operator must select wavelength, pulse width, range, averaging time, launch fiber, receive fiber, and event-detection thresholds. Launch and receive fibers permit measurement of the first and last connector, rather than hiding them within the OTDR's instrument dead zone.

4.3 Common interpretation errors

OTDR distance is not necessarily cable route distance because fiber may have a different refractive index from the value entered in the instrument. A reflective event followed by a temporarily elevated trace can create a measurement dead zone. Two close connectors may be reported as one event. Macrobending may be more visible at 1550 nm than at 1310 nm because bend loss is often wavelength dependent. A two-way or bidirectional test is preferable for accurate splice-loss evaluation, because differential backscatter coefficients can create apparent gain or loss in a one-way trace.

OTDR does not directly measure end-to-end insertion loss with the same meaning as a calibrated source-and-meter test. It estimates distributed and event losses through backscatter analysis. Consequently, acceptance testing commonly uses both an insertion-loss test and OTDR documentation.

5. BER Testing and Eye-Diagram Analysis

5.1 BER and confidence

BER is defined as

BER = Ne/Nb,

where Ne is the number of erroneous bits and Nb is the number of tested bits. A measured zero-error result is not proof that the true BER is zero. If errors follow an independent Bernoulli model and zero errors are observed in N trials, the probability of zero errors when the true BER is p is (1-p)^N. Thus, confidence statements require a specified statistical model and test length.

For small p, a common approximate planning rule is that observing zero errors over N bits gives approximately 95% confidence that p is below about 3/N. This is an approximation based on the Poisson limit and should not replace a formal confidence calculation when standards or contracts require one.

A BER tester (BERT) generates a known pattern, applies it to the transmitter or channel, recovers the received data, aligns the pattern, and counts errors. Patterns such as pseudorandom binary sequences (PRBS) exercise transitions and stress equalization differently. The selected pattern, data rate, coding, test duration, clock source, and error threshold must be recorded.

5.2 Eye diagrams

An eye diagram superimposes many symbol intervals after triggering on the data clock. Vertical closure indicates amplitude noise, intersymbol interference, extinction-ratio degradation, or receiver bandwidth limitations. Horizontal closure indicates timing jitter, duty-cycle distortion, and dispersion-induced pulse spreading. Eye height and width are useful diagnostic indicators, but compliance requires the measurement method, filter, sampling bandwidth, reference clock, and mask to be specified.

For high-speed systems, an eye may be measured electrically at a transmitter or receiver, or optically using an optical sampling system. An optical eye measurement and an electrical eye measurement are not interchangeable. Eye diagrams are excellent for locating impairment mechanisms but do not replace BER testing, because a mask margin is not identical to an error probability.

5.3 Relationship among power, eye, and BER

Power measurement answers whether sufficient average optical energy is present. The eye diagram shows waveform quality and timing behavior. BER directly evaluates data decisions under a specified test pattern and receiver. A robust diagnostic sequence uses all three: low power suggests attenuation or transmitter failure; a closed eye at adequate power suggests dispersion, noise, bandwidth limitation, or bias problems; high BER with a visually open eye may indicate clock recovery, pattern synchronization, intermittent faults, or an unsuitable measurement setup.

6. Commissioning and Acceptance Testing

Commissioning should be planned as a controlled process rather than a collection of ad hoc readings. A useful sequence is:

1. Review design documents, link budget, equipment specifications, fiber route, connector plan, and applicable standards.

2. Verify instrument identity, calibration status, operating range, software version, and environmental conditions.

3. Inspect cable labels, bend radius, grounding, rack installation, connector end faces, and optical safety controls.

4. Perform continuity and polarity checks, followed by visual inspection and cleaning.

5. Measure fiber attenuation or insertion loss at relevant wavelengths using a calibrated source and meter.

6. Acquire OTDR traces in both directions where splice or event-loss analysis is required.

7. Measure transmitter and receiver powers, amplifier gain, optical return loss, and WDM spectra as applicable.

8. Activate the data path and test BER, eye quality, synchronization, alarms, protection switching, and management functions.

9. Compare results with contractual limits, equipment specifications, and the design budget.

10. Record raw traces, instrument settings, calibration certificates, environmental conditions, operator identity, and corrective actions.

Acceptance limits should be defined before testing. A result should not be declared acceptable merely because it is close to a nominal value. Measurement uncertainty and guard bands are important when a result lies near a specification boundary.

7. Calibration, Uncertainty, and Data Integrity

Calibration compares an instrument with a reference of known value and documents the relationship, correction, and uncertainty. Calibration is not the same as adjustment. Adjustment changes the instrument; calibration characterizes it. Instruments should be calibrated at intervals justified by stability, usage, manufacturer guidance, regulatory requirements, and historical drift.

A measurement result should ideally be reported as an estimate together with an uncertainty statement. Sources include detector calibration, wavelength mismatch, connector repeatability, source stability, meter resolution, temperature, polarization, and operator technique. If independent standard uncertainties are combined and a linear model is appropriate, the combined standard uncertainty may be estimated by the root-sum-square method:

uc = sqrt(u1^2 + u2^2 + ... + un^2).

This expression assumes the component uncertainties are appropriately expressed in the same units and are uncorrelated. Correlated components require covariance terms. An expanded uncertainty U = kuc uses a stated coverage factor k; k = 2 is often used for an approximately 95% coverage probability when the statistical assumptions are reasonable.

A calibration system should preserve traceability to a national metrology institute or accredited laboratory. Digital records should be protected against unauthorized alteration, with version control, time synchronization, access control, and retention policies. Raw traces are evidence; edited screenshots without metadata are weak evidence.

8. Fault Diagnosis and Maintenance

Fault diagnosis is most effective when hypotheses are tested from the physical layer upward. A disciplined process is:

• Confirm the symptom and establish whether it is permanent, intermittent, wavelength-specific, direction-specific, or load-dependent.
• Check alarms, recent changes, patching records, and environmental history.
• Verify instrument operation using a known-good reference.
• Inspect and clean connectors before replacing components.
• Measure transmit and receive powers to localize a loss of optical energy.
• Use OTDR from both ends to locate reflective events, breaks, bends, and abnormal splice regions.
• Use the OSA to investigate wavelength drift, channel loss, ASE, or crosstalk.
• Use eye and BER measurements to distinguish waveform impairment from protocol or clock problems.
• Substitute known-good components only after measurements support the hypothesis.
• Document the root cause, corrective action, verification test, and recurrence-prevention measure.

Typical fault signatures include a sudden power drop caused by a disconnected or contaminated connector; increased attenuation at 1550 nm caused by a bend; a strong OTDR reflection caused by an open connector; multiple WDM channels degraded together because of a common amplifier or fiber fault; and a single channel degraded because of a transceiver, filter, or wavelength issue.

Preventive maintenance includes connector inspection, cleaning, patch-cord management, environmental monitoring, calibration review, alarm testing, software and configuration backup, and trend analysis. Repeated measurements can reveal gradual power decline before service failure.

9. Reliability Engineering of Optical Links

Reliability is the probability that a system performs its required function for a specified time under stated conditions. Availability additionally includes restoration time and logistics. A simple series model illustrates the importance of each subsystem: if independent elements have reliabilities R1, R2, ..., Rn over the mission interval, system reliability is

Rsystem = product from i=1 to n of Ri.

The independence assumption is often unrealistic because common causes—construction damage, power failure, temperature, or software defects—can affect several elements simultaneously. Redundant paths, diverse routing, automatic protection switching, spare modules, and rapid repair improve availability, but add cost and may introduce switching complexity.

Failure-rate models must be used cautiously. The constant failure-rate exponential model is useful for limited periods when a component's hazard rate is approximately constant:

R(t) = exp(-lambda t),

where lambda is the constant failure rate and t is time. Real optical components may exhibit infant mortality, a useful-life region, and wear-out behavior; therefore accelerated testing, field-return analysis, and condition monitoring are needed for credible predictions.

Reliability-centered maintenance asks what functions can fail, what causes each failure, what consequences result, and what preventive or predictive task is technically justified. Excessive testing can itself create connector wear, service interruption, or contamination. Maintenance policy should balance risk, measurement uncertainty, and operational cost.

10. Standards and Governance

Standards provide common definitions, test methods, interfaces, safety requirements, and performance expectations. Relevant families include:

• ITU-T recommendations: Optical interfaces and transport-system characteristics, including fiber and cable characteristics, optical network performance, and digital transmission requirements. The applicable recommendation depends on network type and interface.
• IEEE standards: Ethernet physical-layer specifications, including optical media, data rates, reach, coding, and link requirements. The exact IEEE 802.3 clause must be identified rather than citing “IEEE” generically.
• IEC standards: Fiber-optic component and measurement procedures, calibration-related practices, environmental tests, connector and cable requirements, and safety-related requirements. IEC 61280 test procedures are particularly relevant to fiber-optic communication subsystems, while IEC 60825 addresses laser product safety.
• ISO/IEC 17025: Competence requirements for testing and calibration laboratories, supporting traceable and technically valid results.

Standards evolve. Engineers must consult the current edition, national adoption, amendments, and contractual interpretation. A vendor datasheet is not automatically a standard-compliance certificate. Test reports should name the document, edition, clause or method, equipment configuration, and pass/fail criterion.

Laser safety requires classification awareness, controlled access, warning labels, interlocks where necessary, beam management, and procedures that prevent viewing disconnected or active fiber ends. Electrical safety, electromagnetic compatibility, information security, and privacy of captured network data are also governance concerns. Test systems connected to production networks can expose management interfaces or customer traffic; access control and approved test windows are essential.

11. Applied Case: Illustrative Metropolitan WDM Link

Background

A 40 km metropolitan WDM link exhibits intermittent degradation on one channel after a maintenance activity. The nominal design uses 1550 nm transmission, two passive connector interfaces, several fusion splices, and an optical amplifier.

Problem

The network alarm reports occasional high BER on channel 3, while other channels remain within limits. Average received power on channel 3 is 2 dB below its commissioning value. An initial technician suspects a failed receiver.

Analysis

The maintenance team first verifies the power meter against a reference source and inspects and cleans all accessible connectors. Channel 3 remains low, while the other channels are unchanged. The OSA shows that the channel center wavelength is correct, but its power is reduced and a small amount of adjacent-channel leakage is present. This makes a channel-specific optical path problem more likely than a common amplifier failure.

Bidirectional OTDR tests are performed at 1310 and 1550 nm using launch and receive fibers. A non-reflective loss event is found approximately 18 km from the first terminal. The event is substantially worse at 1550 nm, a signature consistent with a macrobend. The measured distance is corrected using the fiber group index and route records. Inspection finds that a newly installed closure has forced the fiber below its specified bend radius.

Approach and solution

The fiber is rerouted within the closure, the bend-radius requirement is restored, and the closure is resealed. Insertion loss is remeasured using the source-and-meter reference method. The OTDR trace is repeated in both directions. Channel power returns within the commissioning tolerance, and a BER test using the prescribed PRBS pattern records no errors over the required test interval. The eye diagram shows restored horizontal opening.

Outcome

The fault was not a receiver failure. It was a wavelength-sensitive physical installation defect that reduced power and degraded waveform margin on one channel. The final record includes before-and-after power values, OSA spectra, bidirectional OTDR traces, BER settings and results, connector inspection records, and the corrective-action report.

Learning points

The case demonstrates why a single instrument is insufficient. Power located the symptom, the OSA established channel specificity, OTDR localized the physical event, and BER and eye measurements verified service quality. It also demonstrates the value of trend records and maintenance procedures that control bend radius.

12. Contemporary Developments and Best Practices

Automated test platforms increasingly combine power meters, OSAs, OTDRs, traffic generators, and network-management interfaces. Data analytics can compare current traces with commissioning baselines and flag statistically significant drift. Machine-learning models may classify OTDR signatures or predict failures from historical alarms, but they require representative labeled data and careful control of false positives. An algorithm should support, not replace, traceable measurements and engineering judgment.

Coherent systems require additional measurements such as optical signal-to-noise ratio, constellation quality, polarization-related impairments, phase noise, and generalized mutual information. High-speed electrical standards increasingly use statistical compliance methods, calibrated fixtures, de-embedding, and reference equalization. These developments make measurement configuration as important as the displayed result.

Best practice is to define a measurement control plan before installation; use calibrated and suitable instruments; clean and inspect interfaces; preserve raw data; measure at the correct wavelength and operating state; combine complementary instruments; quantify uncertainty; apply current standards; protect personnel and network data; and close every fault with an independent verification test.

Challenges and Limitations

No instrument observes every aspect of a link. Power meters provide limited temporal and spectral information. OSAs may trade resolution for speed and cannot prove BER compliance. OTDR traces are affected by dead zones, backscatter differences, pulse settings, and refractive-index assumptions. BERTs can produce misleadingly favorable results if the pattern, duration, clock, or receiver configuration is inappropriate. Eye diagrams depend strongly on bandwidth, filtering, trigger quality, and sampling methodology.

Field conditions also complicate measurement: inaccessible closures, live traffic, temperature variation, vibration, contamination, uncertain documentation, and incompatible connector types. Acceptance criteria may be ambiguous when standards, vendor specifications, and contract requirements differ. These limitations reinforce the need for documented methods, uncertainty-aware decisions, and independent review of borderline results.

Chapter Summary

Optical communication measurement establishes whether a link meets its design and operational requirements. Optical power meters and calibrated sources measure absolute power and insertion loss; OSAs verify wavelength, channel power, spectral occupancy, ASE, and crosstalk; OTDRs locate events and characterize distributed fiber behavior; BERTs quantify data errors; and eye diagrams reveal amplitude and timing impairments. Each instrument has distinct assumptions and limitations, so reliable diagnosis requires complementary measurements.

Commissioning should combine inspection, calibrated loss testing, OTDR traces, power and spectrum verification, BER testing, and documented acceptance criteria. Calibration, traceability, uncertainty, guard bands, data integrity, and current standards are essential to defensible results. Reliability engineering extends testing into maintenance, redundancy, failure analysis, and availability planning. ITU-T, IEEE, IEC, ISO/IEC 17025, and laser-safety requirements provide an interoperable governance framework. Finally, automated analytics and AI can improve detection and prediction, but only when supported by valid data, traceable measurement, and expert oversight.

Key Terms

Optical power meter: Instrument that measures optical power using a calibrated photodetector.

Insertion loss: Power reduction introduced by a component or link.

Return loss: Ratio in decibels of forward power to reflected power.

Optical spectrum analyzer: Instrument that measures optical power versus wavelength or frequency.

Resolution bandwidth: Effective filter bandwidth determining spectral resolution.

OTDR: Instrument that uses backscatter and reflections to characterize fiber versus distance.

Rayleigh backscatter: Distributed light scattering used by an OTDR for distance and loss analysis.

Dead zone: Distance region near an OTDR event where subsequent events cannot be resolved accurately.

BER: Ratio of erroneous bits to total tested bits.

BERT: Instrument that generates, receives, and counts errors in a known data pattern.

PRBS: Pseudorandom binary sequence used to stress and test digital links.

Eye diagram: Superimposed symbol waveforms used to assess noise, timing, and intersymbol interference.

Traceability: Connection of a measurement through documented calibrations to recognized standards.

Measurement uncertainty: Quantified doubt associated with a measurement result.

Reliability: Probability of performing a required function for a stated time and condition.

Availability: Proportion of time a service is operational, including restoration effects.

Common-cause failure: Failure mechanism affecting multiple supposedly independent elements.

WDM: Transmission of multiple optical wavelengths through one fiber.

ASE: Amplified spontaneous emission noise produced by optical amplifiers.

Guard band: Decision margin used when comparing an uncertain result with a specification limit.

Review Questions

Short Answer Questions

1. Distinguish insertion loss from return loss.

2. Why are launch and receive fibers used in OTDR testing?

3. What does an optical spectrum analyzer measure?

4. Why does a zero-error BER test not prove that the true BER is zero?

5. State two limitations of eye-diagram measurements.

Long Answer Questions

1. Explain the operating principles, measurement parameters, and limitations of an OTDR.

2. Develop a complete commissioning procedure for a point-to-point optical link.

3. Discuss calibration, traceability, uncertainty, and guard bands in optical measurements.

4. Compare the diagnostic roles of a power meter, OSA, OTDR, BERT, and eye-diagram instrument.

5. Explain how reliability engineering and maintenance planning improve optical-network availability.

Analytical / Application Questions

1. A transmitter launches +1 dBm into a 25 km fiber with attenuation of 0.24 dB/km. Connector and splice losses total 2.0 dB. Calculate the estimated receiver power before engineering margin is applied, and state what additional information is needed to assess compliance.

2. An OTDR shows a loss event at an apparent distance of 12 km, but the entered group index is too low. Explain qualitatively how the reported distance is affected and identify a procedure for correcting it.

3. A WDM channel has adequate average power but high BER, while all neighboring channels are normal. Propose a measurement sequence using an OSA, eye instrument, and BERT to distinguish wavelength, waveform, and receiver-clock causes.

Multiple Choice Questions with Answers

Question 1

Which instrument is primarily used to measure optical power as a function of wavelength?

A. Optical spectrum analyzer

B. Optical power meter

C. OTDR

D. BER tester

Correct Answer: A. Optical spectrum analyzer

Question 2

A link has a reference power of -2 dBm and a measured output of -5.5 dBm. What is its insertion loss?

A. 2.0 dB

B. 3.5 dB

C. 5.5 dB

D. 7.5 dB

Correct Answer: B. 3.5 dB

Question 3

In the OTDR distance equation z = ct/(2n), the factor 2 represents:

A. Two wavelengths in the fiber

B. Two detector channels

C. The forward and return propagation paths

D. Two splice losses

Correct Answer: C. The forward and return propagation paths

Question 4

Which OTDR parameter generally improves range but worsens spatial resolution when increased?

A. Pulse width

B. Connector cleanliness

C. Group index accuracy

D. Wavelength label

Correct Answer: A. Pulse width

Question 5

A measured BER is calculated as:

A. Total bits divided by erroneous bits

B. Erroneous bits divided by total tested bits

C. Received power divided by transmitted power

D. Optical loss divided by distance

Correct Answer: B. Erroneous bits divided by total tested bits

Question 6

Vertical closure of an eye diagram most directly indicates reduced:

A. Wavelength accuracy

B. Amplitude margin

C. Fiber route distance

D. OTDR dynamic range

Correct Answer: B. Amplitude margin

Question 7

Which observation most strongly suggests a wavelength-selective physical bend?

A. Equal power reduction at all wavelengths

B. Greater loss at 1550 nm than at 1310 nm

C. A higher BER only when the BERT is disconnected

D. An unchanged OTDR trace at every wavelength

Correct Answer: B. Greater loss at 1550 nm than at 1310 nm

Question 8

The principal purpose of calibration traceability is to:

A. Increase the source launch power

B. Relate measurements through documented calibrations to recognized standards

C. Eliminate all measurement uncertainty

D. Replace acceptance criteria

Correct Answer: B. Relate measurements through documented calibrations to recognized standards

Question 9

For independent uncertainty components, the root-sum-square method is used to estimate:

A. Combined standard uncertainty

B. Optical return loss

C. Fiber group index

D. Bit rate

Correct Answer: A. Combined standard uncertainty

Question 10

Which standard family is most directly associated with Ethernet physical-layer specifications?

A. IEEE 802.3

B. IEC 60825 only

C. ISO 9001 only

D. ITU-T G.711 only

Correct Answer: A. IEEE 802.3