Performance and Energy Trade-Off Analysis of Hierarchical Federated Learning for Plant Disease Classification

This document aggregates 11 recent arXiv submissions spanning distributed AI, edge/IoT systems, and distributed/satellite consensus. Key security-relevant contributions: (1) a Privacy-Preserving Federated Learning architecture that integrates Zero-Knowledge Proofs (ZKPs) to cryptographically validate node computations and mitigate model-poisoning without revealing gradients; (2) OrbitBFT, a hierarchical Byzantine Fault-Tolerant consensus design tailored to LEO satellite constellations (dynamic topology, sparse connectivity) with bypass/hop-by-hop mechanisms to tolerate adversarial behavior; (3