An Efficient Task Offloading Strategy in Cloud-Edge Computing Under Deadline Constraints

Cloud-edge computing is a new paradigm that can offer sufficient resources and low latency for mobile users, by means of offloading computation tasks from mobile devices to cloud/edge servers. This paper aims at scheduling security-critical tasks in cloud-edge environments by proposing a novel firefly algorithm. The optimization goal of the scheduling problem is to minimize the energy consumption of the cloud-edge system under a deadline constraint upon task completion time. The proposed algorithm uses a position-based mapping method to map a firefly onto a high-quality solution depending on the best solution and a probability model. In addition, we develop a linear movement scheme to replace the standard movement scheme for the purpose of efficient update of firefly positions. Experimental results show that, compared with the standard firefly algorithm, our proposed algorithm achieves not only higher scheduling effectiveness but also lower computational complexity.