Decentralized IoT (DIoT) and wireless sensor networks face critical challenges in energy efficiency, resource management, and reliable data transmission. This paper proposes two novel routing techniques that integrate Dynamic Physically Unclonable Function (DPUF) and Grey Wolf Coot Optimization (GWCO) to address these issues. A dynamic fitness function is introduced, incorporating residual energy, traffic load, buffer capacity, hop count, and transmission distance to identify optimal routing paths. Unlike conventional approaches, the proposed methods achieve balanced resource utilization, enhanced fault tolerance, and improved scalability in DIoT environments. Extensive simulations validate their effectiveness, showing significant improvements: communication delay reduced by up to 333 s, energy consumption decreased by 2.97 mJ, packet delivery ratio (PDR) increased to 99.98%, and throughput enhanced to 312 Mbps. These results demonstrate that the GWCO-DPUF framework outperforms traditional routing protocols, offering an energy-efficient, secure, and resilient solution for next-generation decentralized IoT systems.