Table 3 Experimental setup of simulation model.
Component | Specification/configuration |
|---|---|
Hardware platform | NVIDIA RTX 4070 Ti GPU, 32 GB DRAM, Intel i5-13400 CPU with 10 cores at 2.50 GHz |
Software environment | Python 3.10.11, PyTorch 2.0.1, Go HTTP client, Tornado HTTP server |
Operating system | Microsoft Windows 10 professional (x64) |
CPU, memory, graphics card | NVIDIA RTX 4070 Ti GPU, 32 GB DRAM, Intel i5-13400 CPU with 10 cores at 2.50 GHz |
Federated learning framework | Flower (FLwr) for simulation and model distribution |
Blockchain layer | Custom implementation using cryptographic SHA-256 and smart contract rules |
Optimization algorithm | Chaotic Bobcat Optimization Algorithm (CBOA) |
Model type | Entropy deep belief network (EDBN) |
Privacy layer | RAPPOR-based local differential privacy (LDP) |
Security mechanism | SHA-256 for hash computation and blockchain immutability |
Consensus algorithm | Proof of contribution (PoC) |
Train-test split | 20% testing and 80% training |
Metrics for evaluation | Accuracy, precision, recall, F-measure, loss, latency, and throughput |
Attack models | Simulated DYN-OPT, STAT-OPT, label-flipping attack, and additional noise attack |
Technology | Blockchain framework, federated learning |
