H·K·N
Chapter 02 of 07

Research.

What I work on, how it fits together, and what I am building right now.

My research lives at a single intersection: time and structure. On one side, the time-sensitive networking stack — schedules, synchronization, deterministic forwarding. On the other, the algebraic structure of network coding — packets as vectors over a finite field, recovery as solving a linear system. Around both, a wireless edge — 5G, soon 6G — that has to behave like a deterministic fabric without giving up the things that made it cellular.

What ties the threads together is one question: what is the smallest set of guarantees a network has to provide so that a machine on the other end can act on what it sees?

Interests

Current research

Project 2024 — present active

CampusTwin4In

TSN controller design and 6G–TSN control plane integration for a campus-scale industrial digital twin. I focus on the controller side: configuration, scheduling, and the semantics of handing a flow between a 6G slice and a TSN island without losing the guarantees that made it useful in the first place. The work spans the standards (IEEE 802.1Q family), the kernel, and the radio.

stack TSN switches OMNeT++ Linux PTP / gPTP 5G modems CNC controllers
papers
Theme ongoing active

Packet recovery for random linear network coding

Reconstruction strategies that preserve in-order delivery and freshness — including dynamic history length, partial-recovery acceleration, and hybrid recovery with HMAC verification. The goal is reliability that does not cost the latency budget it was meant to protect, and verification that scales without forcing every node to do the same work twice.

stack RLNC over 𝔽2m Homomorphic MAC C / kernel modules measurement testbeds
papers
Theme ongoing active

5G–TSN for edge-controlled robotics

Time-slot-aware transmission in 5G TDD, paired with TSN-side shaping, to make remote closed-loop control feasible over public cellular. Less about peak throughput, more about how predictably a control packet survives the radio. The inverted-pendulum demo over 5G-TSN is the reference workload — when balance holds end to end, the protocol stack is doing its job.

stack 5G testbed TSN bridges OMNeT++ real-time kernels edge controllers
papers
Tooling ongoing active

TSN simulation & measurement testbeds

Improving TSN simulation accuracy in OMNeT++ and contributing to flexible measurement testbeds — including the published Spot-robot tactile-internet traffic dataset on IEEE Dataport. Simulators that lie to you are worse than no simulator at all; the work here is making the numbers honest so anyone else's results are reproducible too.

stack OMNeT++ / INET hardware TSN switches IEEE Dataport PCAP analytics
papers

Demos & lab notes.

Short writeups of things I have built, broken, or measured in the lab. Most of these will grow into proper notes over time — this is where they will live.

note in progress
5G-TSN edge robotics URLLC

Balancing an inverted pendulum over 5G-TSN

Closed-loop control across a public 5G channel and a TSN bridge, with end-to-end prioritization. When the pendulum stays up, the stack is honest. When it falls, you have a very legible bug report.

read note → 2026 · CCNC
note in progress
tactile internet 5G-TSN FlexTac

Touching things across a network

The FlexTac setup: haptic interactions traversing a 5G-TSN link with end-to-end shaping. A demonstration that the tactile internet is less about bandwidth and more about jitter you can trust.

read note → 2024 · INFOCOM
note in progress
network coding RLNC recovery

Fly-PRAC: recovery on the wing

Walking through the random linear network coding recovery path from a real packet trace — what gets reconstructed, what does not, and why the history length is the parameter nobody warns you about.

read note → 2026 · arXiv
note in progress
TSN testbed methodology

TSN-FlexTest in practice

A walkthrough of the measurement testbed: what we measure, where commodity hardware lies to you, and the published traffic-stream dataset for Spot-robot, tactile-internet, and generic flows.

read note → 2023 · IEEE TNSM