Online Algorithm for Secure Task Offloading in Dynamic Networks
Yuchong Luo, Jigang Wu, Zikai Zhang, Wenjun Shi, Yuqing Miu · 2017
With the rapid development of mobile applications, it is a new challenge on computational resource and energy of mobile devices. To guarantee the efficiency of the mobile application under the limited resource, the power consumption and execution time of mobile devices are decreased through offloading a part of tasks to the cloud in mobile cloud computing. Most of existing strategies for multi-objective task offloading are based on static network and it is hard to directly apply them to real dynamic networks. As the encryption and decryption in the uploading and downloading of sensitive information data, this paper proposes a task secure offloading model, followed by an online algorithm for secure task-offloading, utilizing a receding horizon online algorithm to get optimal strategy for minimizing energy and time consumption in mobile cloud computing. The proposed algorithm focuses on the reliability of applications and the energy consumption of communication, and constructs the task reduction graph based on task reliability level, resulting in an online algorithm using pareto-optimal strategy. The robustness of the proposed algorithm on dynamic networks is successfully improved in secure task offloading. Experimental results show that, the real-time solution generated by the online algorithm is very close to the exact solutions on the networks with given messages. In addition, 9% approximate solutions are identical to the corresponding exact solutions. The difference of energy consumption between approximate solutions and exact ones is no more than 20%