Independent research · simulation-validated · seeking evaluation partners

Communication that bends instead of breaking.

A research program on neural joint source-channel coding — radio links where the transmitter and receiver share a learned model, so the signal degrades gracefully and keeps delivering in exactly the conditions that make classical links fail completely: deep-space low-SNR, fading, and jamming.

Honest by design: this page leads with what the technology does and does not do. Every number below is measured in our own simulations, reproducible from the open code.

The core idea

Classical radio works perfectly — until it doesn't.

Today's systems separate the job: compress the data to bits, then wrap those bits in error-correction. Claude Shannon proved that's optimal for infinitely long messages — but it has a hard edge. Above a channel-quality threshold you get a perfect signal; a hair below it, the code fails and you get nothing. That "cliff" is worst exactly where it matters most: at the edge of deep space, through fading, under a jammer.

Image / data Neural encoder
(the shared model)
Noisy / jammed
channel
Neural decoder
(same shared model)
Reconstruction

When both ends share a jointly-trained model, they transmit meaning rather than fragile bits. The result has no cliff: as the channel worsens the reconstruction just gets softer, and every frame still arrives usable. That is the whole thesis — and the sections below show where it holds up and where it doesn't.

The evidence — measured, reproducible

What it demonstrably does

newest · attacks the field's #1 deployment blocker

Rosetta — cross-model interoperability, solved

The single reason semantic communication isn't deployed: two independently-built endpoints learn private codes and can't understand each other (link collapses to noise even on a perfect channel). Rosetta makes three independently-trained radios fully interoperable9.2 → 22.2 dB cross-vendor, every sender-to-receiver pair — via a tiny standardized adapter, with zero changes to the base models. This is the piece a standard (Next G Alliance / 3GPP) would publish.

And the underlying link performance — trained from scratch, tested on standard channel models and real Sentinel-2 satellite imagery against a deliberately strong classical baseline:

+8.8 dB

Works with no feedback

On a one-way link (deep-space, broadcast, jammed) digital must fix its rate and delivers 0% of frames below its design point. The neural link needs zero transmitter channel knowledge and still delivers a usable image — +8.8 dB at 0 dB SNR.

verified · Sentinel-2
0% outage

Every frame arrives

Under Rayleigh fading + barrage jamming, the classical baseline loses ~19% of frames entirely; the learned link delivers 100% at consistent quality, and a hybrid design stays within ~1.3 dB of the best classical system across the whole SNR range without ever cliffing.

verified · hybrid
+3.0 dB

Jamming-hardened

A reactive follower-jammer cuts an unhardened link by 6.3 dB; adversarial training recovers 3.0 dB and the robustness transfers to the worst-case jammer, for <1 dB clean-channel cost.

verified · adversarial
Straight talk

Where it fits — and where it doesn't

Strong fit

RegimeWhy
Deep-space / low SNRgraceful, no cliff, no outage
Contested / jammed (EW)hardens against reactive jammers
Fading, no retransmitdelivers every frame usably
Bandwidth-starved links~12× source compression built in

Poor / unproven fit

RegimeWhy
High-SNR / clean linksrate-adaptive digital wins — hand off
Exact / lossless dataneural decoders can hallucinate
Deployed RF hardwarenot yet tested — simulation only
Interoperabilityboth ends need the same model
Honest limitations — read this first

What this is not (yet)

The full record — nothing hidden

All research & code

Seven studies, each with a full report and open source. They include the failures: an architecture that partly worked, and an adaptivity idea that didn't. That's the point — this is a real research trail, not a brochure.

Rosetta — interoperability

newest · the #1 blocker

Makes independently-trained transceivers understand each other through a shared reference standard + tiny frozen-base adapters. Cross-vendor 9.2 → 22.2 dB, all pairs. The path from lab to standard.

Latent Radio

flagship · LIVE

The core demo: a neural transceiver that beats classical coding across every SNR and degrades gracefully. Interactive — pick or draw an image and dial the noise.

Making it better — hybrid + moat

latest · dominance

Two builds: a hybrid digital-analog code that recovers ~half the good-channel gap without ever cliffing, and the feedback-free moat — usable data on one-way links where digital delivers nothing.

Contested Channel v2

satellite · rigorous

Comms-grade re-test on real Sentinel-2 imagery: Rayleigh/Rician fading, partial-band jammer, honest outage-based baseline. Wins in the hard regime; delivers every frame.

Contested Channel v3

jamming · honest mixed

Reactive/adversarial jamming + adversarial-training defense (worked); SNR-rate-adaptivity (didn't close the gap — reported straight).

Sigil

visual code

A message code that fades instead of breaking — the same principle applied to a QR-like visual channel. 93% bit-recovery at 70% damage.

Aether

emergent language

Two agents invent a private, compressed, noise-robust language to cooperate — cooperation on ~8 numbers per message instead of verbose text.

Contested Channel v1

first stress test

The initial space/defense framing: shared-model transceiver under burst jamming and fades. Motivated the rigorous v2 re-test.

HRM — Harmonic Resonance Mixer

architecture

A from-scratch novel sequence architecture (resonant memory + selective gate). Beats a GRU on noisy frequency-ID; loses on delayed recall — ablation-validated.

Open source

All code, reproducible from scratch.

For evaluators & partners

What would make this real

This is a proof-of-concept with a genuine, narrow advantage and honest gaps. The next steps need resources a laptop can't provide — and that's exactly the conversation we're looking for.

1 · Independent evaluation

Re-run the studies against your own channel models, datasets, and classical baselines. Everything is open and reproducible.

2 · Hardware-in-the-loop

Take it from simulation to an SDR / RF testbed — real Doppler, timing recovery, and front-end nonlinearity — to measure true TRL.

3 · Mission-relevant data

Train and test on representative payloads (hyperspectral, SAR, telemetry) and real link budgets for a target mission profile.

Talk to us

Independent evaluation, testbed access, or a mission-specific study.

priyankar@pchakconsulting.com

Website crafted by Everbuilt Studio