Mobile-aware placement: optimizing energy consumption and delay of data stream applications in edge computing

With the rapid development of 5G and IoT, traditional cloud computing faces limitations due to increasing demands. Edge computing addresses these issues by offloading tasks to nearby nodes, thereby improving efficiency and user experience for latency-sensitive applications such as live video and autonomous driving. However, balancing energy consumption and latency in heterogeneous edge devices remains a challenge, particularly as previous studies often overlook user mobility. In this work, we tackle the problem of data stream processing (DSP) task placement within dynamic user mobility scenarios, addressing the challenges posed by frequent movement. Additionally, we introduce a reinforcement learning approach to enhance the adaptability of DSP tasks, allowing the system to effectively adjust to changing user mobility patterns. Through simulation experiments, it is verified that compared with the traditional method, the algorithm proposed in this paper effectively optimizes energy consumption and delay while considering the mobility attributes, and improves the utilization rate of each edge node.