🌌 QASWP — Quantum-AI Secure Weaving Protocol

---

Overview

QASWP is a hybrid Quantum Key Distribution (QKD) + Neural Semantic Transport framework for the Quantum Internet, achieving predictive compression, verifiable AI inference, and context synchronization between distributed agents.

It fuses quantum security primitives with neural-semantic prediction to create a verifiable, low-latency Layer 4 protocol connecting quantum computers and AI systems.

---

Demo Mode

The demo-only path demonstrates ≥ 99 % semantic compression on repeated templated flows via batched confirmation bits.
It uses a shared-key handshake for testability, a deterministic entanglement sync stub, and a succinct “zk-like” proof (commitment + verify).
Real deployments will vary and require production-grade crypto/ML systems.

---

📈 Claims & Validation

CI: The Proofs workflow in /proofs/ runs reproducible validation tests:

• Compression benchmark: templated flows ≥ 95 % savings (typically ≥ 99 %)
• Shared-session keys: encryption/decryption symmetry confirmed
• Entanglement stub: identical IDs on both ends (0-byte sync)
• Succinct verify: 64-byte proof check passes

---

🚀 Quick Demo (30 seconds)

# 1. Install dependencies
pip install -r requirements.txt

# 2. Run the server in one terminal
python examples/server.py

# 3. Run the client in another terminal
python examples/client.py

Expected Output:

[QASWP Server] Listening for quantum-neural connections...
[QASWP Client] Handshake successful! QBER=0.00% | Session Key derived.
[QASWP Client] Alice predicts next message: "GET /api/v1/profile"
[QASWP Server] Bob predicts next message:  "GET /api/v1/profile"
[Result] Neural Prediction MATCH! (0 bytes of payload transmitted)
[Stats] Compression Ratio: 99.8% 🚀

---

Packet Semantics & Flushing

• Placeholders: non-flushed packets have flushed=False, wire_len=0, and empty nonce/payload; receivers must treat them as no-ops.
• Stream boundaries: call QASWPSession.flush() before shutdown to emit trailing confirmations.
• Schema version: first byte of encrypted payload is schema_version=1 for forward compatibility.

---

📖 Documentation

• Overview
• API Reference
• WHITEPAPER.md: 20-page technical deep dive
• IETF-DRAFT.txt: RFC-style specification
• API_REFERENCE.md: core logic (QASWPSession(is_client=True))

---

🛠 Core Features

• ✅ Hybrid Security: combines QKD (BB84) for eavesdrop detection with Kyber + Dilithium for post-quantum safety
• ✅ Neural Semantic Compression: TinyLLM prediction → single-bit confirmation
• ✅ Verifiable AI: zk-proof-like inference verification
• ✅ Entanglement Simulation: QuTiP-based shared-state model for zero-latency context sync

---

🔮 Future Roadmap

• Qiskit Integration: transition QKD to real quantum backends (2026)
• Federated LoRA Sync: live on-device model updates for edge nodes
• Formal IETF Submission: evolve draft → RFC proposal for QIRG

---

🔐 Licensing & Commercial Use

Public for research and collaboration, but not open-source.

• Non-commercial research use allowed under the CPUTER Inc. Proprietary License (Royalty-Bearing)
• Commercial use requires a paid license and royalties
• Contact info@cputer.com for licensing

See LICENSE and NOTICE.

---

Qiskit-Backed QKD (v2.2 Scaffold)

Feature-flagged optional handshake using Qiskit (IBM Quantum).

export QASWP_QISKIT=1
pip install qiskit
pytest -q tests/test_qiskit_stub.py
pytest -q tests/test_qiskit_integration_optional.py

In CI, the Qiskit Smoke workflow can be triggered manually; it installs Qiskit best-effort and skips cleanly if unavailable.

Current scaffold returns a deterministic 32-byte demo key. Future revisions will derive real keys from Qiskit circuits (Samplers / Estimators).

---

⚖️ Theoretical Limits & Semantic Compression

In classical information theory, the Shannon limit defines the lower bound for lossless compression, set by source entropy ( H(X) ). QASWP does not violate that law. Its “99 % compression” describes semantic efficiency—reducing transmitted bits when peers share predictive context.

[ \text{Effective compression} = 1 - \frac{H(Δ|Ψ)}{H(M)} ]

When ( H(Δ|Ψ)\to0 ), apparent compression approaches 100 %, but total entropy remains consistent with Shannon’s theorem.

References

• Shannon (1948) A Mathematical Theory of Communication
• Zhang et al. (2023) Semantic Communication Networks
• Nedovodin (2025) QASWP v2.1 Technical Report

---

Applicability & Limitations

QASWP excels in structured, context-rich exchanges (telemetry, control-plane, federated updates) where TinyLLM predictions minimize payloads.
Unstructured or high-entropy streams gain little benefit, prompting fallback to classical compression or full-payload transfer.
Hybrid QKD+PQC encryption, authenticated sequencing, and model rotation mitigate replay, leakage, and disclosure risks.

---

🧪 Testing Matrix

• Local: pytest, coverage
• CI: GitHub Actions (CI, Coverage, Lint, CodeQL)
• Environments: Python 3.10 – 3.11 (tox/nox)

---

🐳 Docker

docker build -t qaswp .
docker run --rm qaswp

---

🧰 Dev Container

Open in VS Code → “Dev Containers: Open Folder in Container…”

---

📦 Releasing

• Release notes via Release Drafter
• Version bump in CHANGELOG.md → git tag v2.3 && git push --tags

---

Join the future of communication 🌌 — Star this repo, open issues, or submit PRs.