Task assignment method based on SDN and improved pigeon-inspired optimization in cloud collaboration environment

With the proliferation of smart devices, data traffic is growing rapidly. How to achieve efficient task offloading under the cloud-mist collaboration architecture has become an urgent problem to be solved. For delay-sensitive application scenarios, a task allocation method based on software defined networking (SDN) and improved pigeon-inspired optimization is proposed. First, taking the lowest delay as the optimization goal, a SDN-based task scheduling model is established. Second, the K-means clustering algorithm is used to cluster the tasks arriving within a period of time according to the two attributes of the minimum delay requirement and the calculation amount of the task, thereby multiple task clusters are generated and set respectively according to the minimum delay requirement priority. Then, an improved hybrid quantum pigeon-inspired optimization (QPIO) is set up to select appropriate nodes for task allocation in order of priority. In the algorithm, the location information of each pigeon is mapped to the allocation strategy for all tasks, and the encoding method is discretized. Finally, the proposed method is verified and evaluated by simulation experiments. The results show that the method has better performance in reducing the total delay compared with other methods.