<p>The rapid growth of IoT devices overcomes traditional cloud architectures, highlighting the need for decentralized solutions, such as fog computing, to address challenges in scalability, latency, and security for efficient real-time sensitive data processing. However, fog computing itself faces challenges, including the need for a scalable infrastructure, minimizing latency, and ensuring robust security in decentralized environments. This article presents a novel approach to optimizing edge-fog-cloud data transfer using advanced smart fragmentation techniques with a hybrid approach of combining lightweight fragmentation at the edge device and heavyweight fragmentation at fog devices to ensure efficient, secure transmission while dynamically adapting to resource constraints and proximity. The fragmentation scheme applies the HashShatter technique on the edge device to securely fragment data and protect it with AES encryption during transmission. At the fog device, the system evaluates the resources of adjacent fog nodes using the Resource-Aware fragmentation technique and trust scores, selects the optimal fog node based on these factors, and transmits the encrypted data accordingly. This adaptive framework optimizes latency, scalability, and security, delivering a robust solution for data-intensive applications in edge-fog-cloud environments. Simulations in healthcare IoT scenarios demonstrate the effectiveness of the proposed approach, with Smart Data Flow achieving 3.46 ms latency, 45.38% improved scalability, and 64.58% enhanced resource utilization, proving it to be more efficient than traditional cloud-based and decentralized processing methodologies.</p>