Key points are not available for this paper at this time.
Mobile users run complex applications increasingly, requiring much more computing power and energy. Cloud-enhanced small cell networks, endowing small cells with cloud computing abilities, extend mobile devices capabilities and battery lifetime by providing them with the advantages of instant and high-rate access to the cloud servers. However, conventional offloading policy can not fully exploit these benefits. In this paper, we present a fine-granularity offloading policy, aiming at minimizing the energy consumption while satisfying a strict delay constraint. We consider a practical application consisting of a set of tasks and model it as a generic graph topology. Then, the energy-efficient task offloading problem is mathematically formulated as a constrained 0-1 programming. To solve the problem with a low computing load, we adopt BPSO algorithm. Simulation results show the proposed offloading policy saves up to 25% energy consumption compared with local execution while delay constraint is satisfied. Moreover, it outperforms conventional rough-granularity offloading policy in energy-saving.
Deng et al. (Sun,) studied this question.