Engineering a Resilient Tier III ↔ Tier IV DC-DC Backbone Using Dark Fiber and DWDM Transport

Building a National-Scale Government Data Center Interconnect with High Availability, Optical Diversity, and Operational Resiliency

Modern government and enterprise digital services increasingly depend on geographically separated data centers operating as a unified infrastructure platform. Disaster recovery replication, cloud onboarding, inter-agency services, and mission-critical workloads require a transport backbone capable of delivering deterministic latency, high throughput, and carrier-grade resiliency.

This article shares the engineering supervision experience from the client side during the implementation of a national-scale government Data Center to Data Center (DC-DC) backbone network interconnecting a Tier IV data center in Gazipur and a Tier III data center in Dhaka using leased dark optical fiber infrastructure and Huawei DWDM transport systems.

The project focused heavily on:

• Physical fiber diversity
• Layer-1 transport resiliency
• Optical budget engineering
• DWDM commissioning and validation
• End-to-end failover testing
• Operational handover readiness

The backbone was designed to support:

• Disaster Recovery (DR) replication
• Future cloud interconnectivity
• Government digital platform hosting
• Enterprise-grade workload mobility
• Scalable 100G optical transport services

Project Overview

Parameter                       Design Specification
---------------------------------------------------------------------
Interconnect Type               DC-DC Backbone
Data Centers                    Tier IV (Gazipur) ↔ Tier III (Dhaka)
Transport Technology            DWDM
Fiber Type                      Single Mode G.652D
Fiber Ownership                 Leased Dark Fiber from NTTN Operators
Protection Model                DWDM Optical Line Protection (OLP)
Transmission Capacity           100G Coherent
Wavelength Capacity             16 Wavelengths (C-Band)
Routing Model                   Pure Layer-2 Transport
Fiber Paths                     Primary + Secondary Diverse Paths
Monitoring Platform             Huawei U2000

Why Dark Fiber + DWDM?

Instead of relying on managed bandwidth services from carriers, the organization selected dark optical fiber infrastructure combined with dedicated DWDM transport equipment.

This provided several strategic advantages:

• Full control over optical transport
• Dedicated wavelength scalability
• Future 100G/400G expansion capability
• Deterministic latency
• Better DR replication performance
• Independent Layer-1 resiliency engineering
• Controlled optical supervision and monitoring

The architecture was intentionally designed to separate the transport layer from the IP layer, ensuring transport independence for future cloud and backbone integrations.

Layer-1 Transport Architecture

The backbone architecture was built around dual physically diverse optical paths between the two data centers.

Fiber Path Design

Primary Path

• Approximate distance: 55 km
• Combination of underground armored fiber
• Metro transmission route

Secondary Path

• Approximate distance: 65 km
• Fully geographically diverse route
• Avoided overlap with primary transmission corridor

One of the major design objectives was ensuring the two optical paths never entered the same underground duct route before reaching the respective Meet-Me-Rooms (MMRs) of the data centers.

This requirement became one of the most operationally challenging aspects of the implementation.

Physical Diversity Topology

1. Physical Fiber Diversity Topology

Fiber Diversity Engineering Challenges

In theory, diverse routing sounds straightforward.

In practice, implementing true physical diversity in developing urban environments is extremely difficult.

Several implementation risks were identified early:

1. Shared Duct Risk

Some NTTN providers unintentionally routed “diverse” fibers through common underground ducts for portions of the path.

This creates a hidden single point of failure.

A single excavation incident could simultaneously impact both primary and secondary circuits.

Mitigation

• Continuous route verification
• Joint field inspection with vendors
• GIS/path validation
• Mandatory route audit before acceptance
• Escalation when common duct sections were identified

2. Road Construction Activities

Frequent urban road excavation activities introduced major operational risks.

In several sections, ongoing construction work threatened fiber integrity during deployment.

Mitigation

• Daily coordination with transmission vendors
• Active supervision during civil work
• Immediate rerouting discussions
• Protection reinforcement for vulnerable segments
• Emergency restoration planning

3. MMR Entry Path Convergence

Even when long-haul routes were diverse, some providers attempted to converge fibers near the final MMR approach.

This was strictly rejected.

The engineering requirement mandated complete diversity until MMR ingress.

Mitigation

• Separate building entry paths
• Independent riser verification
• MMR patch panel segregation
• Physical tracing validation

Optical Budget Planning

Optical budget engineering was one of the most critical parts of the project.

The design target was:

This included:

• Fiber attenuation
• Connector loss
• Fusion splice loss
• Patch panel insertion loss
• MUX/DEMUX attenuation

Optical Budget Considerations

Parameter                       Typical Value
------------------------------------------------
Fiber attenuation               0.22–0.30 dB/km
Fusion splice loss              0.05–0.1 dB
Connector insertion loss        0.2–0.5 dB
MUX/DEMUX insertion loss        Vendor dependent
Safety margin                   Mandatory

The engineering supervision team validated:

• OTDR traces
• Splice quality
• Connector cleanliness
• End-to-end attenuation consistency

Several deployed dark fibers showed degradation beyond initial provider claims due to:

• Aging fiber segments
• Poor historical splicing
• Excessive patching
• Legacy transmission infrastructure reuse

DWDM Platform Deployment

The optical transport backbone used Huawei DWDM infrastructure with coherent 100G transmission capability.

The deployment included:

• Bidirectional Fiber Interface Board
• Optical Line Protection (OLP) Board
• Enhanced 8-Channel Optical Add/Drop Multiplexer
• Optical Booster Board with Optical Power Adjustment
• 11× Multi-rate Optical Transponder Board
• 8-Channel Optical Power Monitor Board
• Four-Directional OSC and Timing Transmission Unit
• Integrated Advanced OTDR

DWDM Layer-1 Architecture

2. DWDM Layer-1 Architecture

Real-World DWDM Deployment Challenges

1. Manual Nominal Gain Adjustment

One of the biggest field challenges involved optical power inconsistency caused by:

• Fiber aging
• Poor splice quality
• Uneven attenuation
• Variable insertion loss

Default vendor-recommended optical gain values did not always achieve acceptable received optical power levels.

Mitigation

• Manual optical gain tuning
• Continuous optical power monitoring
• Multi-stage attenuation validation
• Coordinated tuning between both DWDM nodes

This process required repeated fine adjustments during commissioning.

2. Patch Cord Type Compatibility

Mismatch between:

• MMR patch panels
• DWDM interfaces
• Fiber connectors

created avoidable attenuation and operational complexity.

Critical Validation Areas

• LC/UPC vs LC/APC compatibility
• Patch panel polarity
• Connector cleanliness
• Patch cord bend radius

3. Patch Cord Management

Large-scale patching inside MMR environments became operationally sensitive.

Improper patch cord length planning caused:

• Congestion
• Signal tracing complexity
• Airflow obstruction
• Maintenance difficulties

Mitigation

• Physical route survey
• Predefined patch cord length matrix
• Standardized tagging policy
• Color-coded path identification

This significantly improved operational maintainability.

Optical Validation and Acceptance Testing

The acceptance process was extremely detailed because the backbone would carry mission-critical workloads.

The validation included:

OTDR Validation

• End-to-end fiber trace analysis
• Splice point identification
• Reflection analysis
• Distance verification

Optical Power Validation

• TX/RX power measurement
• Attenuation consistency checks
• Booster output verification

BER Testing

Bit Error Rate testing validated optical transmission stability under load.

Latency Testing

Latency consistency between primary and secondary paths was validated to ensure DR replication performance.

Fiber Polarity Validation

Polarity mismatch checks were conducted before production activation.

End-to-End Wavelength Validation

Every wavelength was validated independently.

Optical Failover and Resiliency Testing

The backbone implemented Optical Line Protection (OLP) switching between primary and secondary transport paths.

Failover Validation Objectives

• Zero manual intervention
• Stable traffic continuity
• Minimal switching delay
• Alarm propagation validation
• Service restoration verification

OLP Failover Workflow

3. OLP Failover Workflow

Routing and Backbone Resiliency

The network intentionally used pure Layer-2 optical transport.

This simplified:

• DR extension
• Storage replication
• VLAN extension
• Future cloud onboarding

The DWDM transport layer absorbed optical failures while upper-layer services remained operational.

The architecture also reduced operational complexity compared to introducing MPLS or overlay transport during the initial phase.

Operational Readiness and Handover

One of the most overlooked areas in infrastructure deployment is operational transition.

A technically successful deployment can still fail operationally without proper handover procedures.

The client-side engineering supervision therefore focused heavily on:

• Documentation accuracy
• Alarm verification
• Fiber labeling
• Patch panel mapping
• OTDR archive collection
• Power baseline recording
• Spare inventory validation

Handover and Acceptance Workflow

4. Handover and Acceptance Workflow

O&M Readiness and Monitoring Strategy

Huawei U2000 was used for centralized monitoring and operational management.

The monitoring scope included:

• Optical power levels
• LOS alarms
• OLP switching status
• Transponder health
• Wavelength performance
• Historical alarm tracking
• OTDR-based fiber fault localization with automatic identification and calculation of the exact physical distance to fiber cuts

The operations team was also trained on:

• Optical troubleshooting
• Fiber isolation procedures
• Power degradation analysis
• Emergency restoration coordination

Lessons Learned from the Field

Several engineering lessons became clear during implementation:

Physical diversity must be validated physically, not contractually

A provider’s “diverse route” statement should never be accepted without field verification.

Optical quality degrades over time

Dark fiber condition varies significantly depending on historical usage and maintenance quality.

MMR planning matters more than expected

Poor patch management creates long-term operational problems.

Optical commissioning is iterative

Real-world DWDM deployment rarely works optimally with default optical values.

Continuous supervision is essential

Field deployment quality improves dramatically when engineering supervision remains active throughout implementation.

Final Thoughts

Building a resilient DC-DC backbone is far more than installing optical equipment between two buildings.

It requires:

• transport engineering,
• optical science,
• physical infrastructure validation,
• operational discipline,
• and long-term resiliency planning.

For mission-critical government and enterprise workloads, the reliability of the transport layer directly impacts business continuity, disaster recovery capability, and national digital infrastructure resilience.

This project demonstrated that combining physically diverse dark fiber infrastructure with carrier-grade DWDM transport can deliver a scalable, resilient, and future-ready backbone capable of supporting modern digital transformation initiatives.

As data centers continue evolving into interconnected cloud ecosystems, resilient optical transport networks will remain one of the most critical foundations beneath the entire digital infrastructure stack.

• Physical tracing validation