Building the quantum
internet, one state
at a time.

Most early proposals for quantum networks lean on Gaussian states because the math is friendly — they're easy to prepare and easy to analyze. But Gaussian states have a hard ceiling: several tasks central to a working quantum internet, from entanglement distillation to certain error-correction routines, are provably impossible with Gaussian resources alone. Non-Gaussian operations — photon subtraction, addition, catalysis — break that ceiling. The cost is that the theory gets harder and the experiments get noisier. Most of my Ph.D. work has lived in that tension: characterizing exactly how much non-Gaussianity buys you, and how much of it survives realistic, lossy channels.

Presenting a poster is a different exercise from writing a paper — you get maybe ninety seconds before someone decides whether to stay. What struck me most at ICPQIQC wasn't any single talk but the shift in the room's center of gravity: five years ago most of the audience at a photonics and quantum information meeting was asking "does this work in theory?" This year, far more people were asking "what does this cost in a real fiber link?" That's a healthy sign for the field — and a reminder that the gap between clean theory and lossy hardware is where the interesting problems now live.

Seawater is a strange medium to send quantum states through — it absorbs, it scatters, and its optical properties shift with depth, salinity, and turbidity. Yet underwater links matter: submarines, ocean sensor networks, and coastal infrastructure all need secure communication that free-space or fiber links can't easily reach. Working through the continuous-variable formalism for this channel, the encouraging result is that CV-QKD's tolerance for loss — the same property that makes it useful over long fiber spans — carries over reasonably well underwater, at least over the depths and distances we modeled. There's real engineering left before this leaves the simulation, but the physics doesn't rule it out.