CavinKrenik/QRES: TensorSwarm: Publication Release v15.2

📄 TensorSwarm: Biologically-Inspired Secure Federated Learning for Edge IoT Complete publication release including academic paper, implementation, benchmarks, and reproducibility artifacts. 📖 Paper Download PDF → Title: TensorSwarm: Biologically-Inspired Secure Federated Learning for Edge IoT Devices Author: Cavin Krenik (Olympic College) Pages: 14 Keywords: Federated Learning, Edge AI, SNNs, Differential Privacy, Secure Aggregation 🔬 Key Contributions Fixed-point SNN ensemble for deterministic temporal prediction (Q16.16 arithmetic) Complete privacy stack: Differential Privacy + Secure Aggregation + Zero-Knowledge Proofs Empirical evaluation: Azure VMs (10-100 nodes), real IoT datasets 📊 Experimental Results Compression: 22:1 ratio on time-series sensor data Privacy overhead: 20-30% for full security stack Byzantine tolerance: Up to 49% malicious nodes (Krum aggregation) Regime change recovery: 3-4 rounds average Scalability: Tested up to 100 Azure B1s VMs 🛠️ Implementation Language: Rust (no_std compatible for bare-metal) Security: ed25519 signatures, ECDH masking, Pedersen commitments Networking: libp2p for P2P swarm coordination Deployment: Docker + Azure VM scripts included 📦 Contents paper/ ├── TensorSwarm__...pdf (Publication PDF) ├── paper.tex (LaTeX source) ├── references.bib (14+ citations) └── figures/ (5 publication-quality figures) qres_rust/ ├── qres_core/ (Rust implementation) ├── qres_daemon/ (P2P coordinator) └── security modules (DP, Secure Agg, ZK) reproducibility/ ├── Dockerfile └── scripts/ 🎯 Target Venues Primary: FLICS 2026 (June 9-12, Valencia, Spain) Categories: cs.DC, cs.LG, cs.CR 📚 Citation @software{krenik2026tensorswarm, author = {Krenik, Cavin}, title = {{TensorSwarm: Biologically-Inspired Secure Federated Learning for Edge IoT Devices}}, month = jan, year = 2026, publisher = {Zenodo}, version = {v15.2-paper}, url = {https://github.com/CavinKrenik/QRES} } 📧 Contact Author: Cavin Krenik Institution: Olympic College, Shelton, WA, USA GitHub: https://github.com/CavinKrenik/QRES 🚀 Ready for submission to FLICS 2026 and archival on Zenodo!