(ICWS 2024) Workflow Task Offloading Upon MEC: A Novel Mobility-Aware and Clustering-Based Approach

The widespread deployment of intelligent Internet of Things (IoT) devices brings strict latency demands on complex workload patterns such as workflows. In such scenarios, tremendous data is generated and processed in accordance with specific service chains. Mobile Edge Computing (MEC) has proven its feasibility in reducing the traffic in the core network and relieving cloud datacenters of fragmented computational demands.However, existing solutions to multi-workflow scheduling and offloading in MEC are still limited due to the fact that they usually make task dispatching decisions prior to real execution, making it difficult to cope with the dynamicity of the environment and the mobility of users.To address this challenge, we developed a novel computation offloading strategy by synthesizing a Harris Hawks Optimization-based particle filter trajectory prediction algorithm for forecasting future user locations, a clustering-based multi-workflow merging algorithm for identifying redundant tasks and a Lyapunov optimization-based DQN algorithm for yielding computation offloading schedules. Simulation results clearly show that method beats traditional ones across multiple performance metrics.