ndn-ndt-sync

NDN-based Network Digital Twin Synchronisation Protocol for O-RAN / 6G

This repository contains the simulation work and protocol design for synchronization between Physical Twin and Network Digital Twin — an MSc research internship project at LIST Luxembourg, supervised by Dr. Ayat Zaki Hindi.

---

Research Summary

Network Digital Twins (NDTs) require continuous, low-latency state synchronisation from physical network nodes to their virtual replicas. Existing solutions use IP-based protocols (gRPC/gNMI, NETCONF/YANG) which suffer from redundant data transfer, lack of caching, and poor mobility support in 6G O-RAN architectures.

This project proposes using Named Data Networking (NDN) as the synchronisation substrate, with synchronisation decisions driven by intent class (URLLC / eMBB / mMTC) and the ODTE (Operational Digital Twin Effectiveness) quality metric.

Key Contribution

A three-tier sync protocol:

Tier	Name	Trigger	Mechanism
1	Predictive Sync	ML-predicted state volatility	Interest sent early; compensates for T_transfer variability
2	Corrective Sync	ODTE trending below threshold	Increased sync frequency for affected data types
3	Emergency Sync	ODTE below threshold	Aggressive Interest rate, MustBeFresh, sensitivity-ranked variable selection

---

Repository Structure

ndn-ndt-sync/
│
├── baseline/
│   └── liang-2023/          ← Replication of Liang et al. (2023) — NDN for DT-IoT
│       ├── liang-small.cc   ← ndnSIM simulation (3×3 grid, Scenario 1)
│       ├── run_all.sh       ← Parameter sweep script (5 BW × 4 rates = 20 runs)
│       ├── parse_results.py ← Parses L3 tracer output → metrics CSV
│       └── README.md        ← Full setup and reproduction guide
│
├── results/
│   └── liang-2023/
│       ├── liang-small-metrics.csv   ← Parsed reception ratio + delay per run
│       └── liang-small-summary.csv  ← Run completion log
│
└── docs/
    └── problem-definition.md

Raw trace files (*-l3-*.txt, *-delay-*.txt) are excluded via .gitignore — they are large and fully reproducible from the simulation source.

---

Key Papers

Paper	Role in this project
Liang et al. (2023) — NDN for DT-IoT Data Collection	Baseline: proves naive NDN sync saturates at ~50% packet drop beyond 30 sub-networks. Motivates INDS selective sync.
Bellavista et al. (2024) — ODTE Metric	ODTE = T × R × A. Dropping below threshold is the trigger for INDS sync tiers.
Kalasapura (2023) — TwinSync	Variable sensitivity logic used in Tier 3 emergency sync.
Sengendo & Granelli (2024) — NDT Sync Delay Chain	Formal delay decomposition; negative delay via ML prediction motivates Tier 1.

---

Status

• Problem definition formalised
• Liang (2023) baseline simulation implemented and validated
• Protocol design (in progress)
• ndnSIM implementation of Tier 1
• Evaluation against gNMI baseline

---

Environment

Simulations run inside the official ndnSIM Docker container. See baseline/liang-2023/README.md for full setup instructions.

# Quick start (after Docker setup)
docker run -it --rm -v $(pwd):/sim ndnsim/ndnsim:latest bash

---

Author

Marium — MSc Cybersecurity & Cyber Defence, University of Luxembourg
Research Internship, LIST Luxembourg