Optimal Energy Efficiency With Delay Constraints for Multi-Layer Cooperative Fog Computing Networks

We develop a joint offloading and resource allocation framework for multi-layer cooperative fog computing network, aiming to minimize the total energy consumption of multiple mobile devices subject their service delay requirements. The resulting optimization involves both binary (offloading decisions) real variables (resource allocations), making it an NP-hard computationally intractable problem. To tackle it, we first propose improved branch-and-bound algorithm (IBBA) that is implemented in centralized manner. However, due large size computational complexity proposed IBBA relatively high. speed up optimal solution searching as well enable its distributed implementation, then leverage unique structure underlying problem parallel processing at nodes. end, framework, namely feasibility finding Benders decomposition (FFBD), decomposes original into master decision subproblems allocation. (MP) equipped with powerful cutting-planes exploit fact limitation (SP) can find closed-form solutions using our fast detection method. These (simpler) be solved numerical results show FFBD always returns significantly less computation time (e.g., compared approach). method, FFBD-F, reduce 60% 90% time, respectively, those conventional FFBD, FFBD-S, IBBA.