Secure Data Transmission and Privacy-Preserving Path Control for Distributed IoT Based on SDN and Mathematical Optimization

INTRODUCTION: Distributed Internet of Things (IoT) communication must jointly satisfy low-latency forwarding, data confidentiality, integrity, access control, and lifecycle protection. Performance-oriented routing may expose sensitive traffic to malicious relays, high-risk links, unauthorized domains, and tampered flow entries. OBJECTIVES: This study reformulates route optimization as a secure data-transmission and privacy-preserving path-control problem. METHODS: A three-layer SDN framework combines robust node-reputation and link-risk scoring, security-label matching, attribute-based access control, flow isolation, AES-256-GCM authenticated encryption with ephemeral keys, signed rules, and a diversity-aware improved genetic algorithm. A composite objective integrates delay, loss, utilization, route risk, privacy exposure, and hard policy constraints. A reproducible Python/NetworkX discrete-event simulation compares Dijkstra, ECMP, AODV, OLSR, RPL, classical GA, SDN-TE, trust-aware routing, and two proposed variants. RESULTS: Across eight independent runs, the full scheme achieved 16.68 ± 1.35 ms normal-operation delay and 98.54 ± 0.21% packet delivery. With 20% malicious nodes, it maintained 98.18 ± 0.39% packet delivery, reduced malicious-link exposure from 20.83% for Dijkstra to 6.67%, eliminated observed plaintext leakage of sensitive payloads in the main runs, and reduced attack-response latency from 722.22 ± 23.53 ms to 130.92 ± 6.34 ms. Paired comparisons against Dijkstra were significant for malicious-link exposure, abnormal-path selection, leakage, packet delivery, and response latency (p < 0.05). CONCLUSION: The framework converts security and privacy requirements into enforceable routing, cryptographic, access-control, isolation, auditing, and lifecycle policies, enabling reproducible security-performance co-optimization for distributed IoT.