Intelligent edge services should consider not only service provider’s resource constraints but also end user’s quality of experience (QoE) demands. In this regard, we study a digital twin service caching and request routing problem with consideration of both resource and fairness-awareness under the collaboration of multi-access edge computing (MEC) servers. We first formulate the problem as an integer nonlinear programming by maximizing the minimum task completion ratio when there are multiple heterogeneous digital twin services, subject to computation, storage, and heterogeneous task delay constraints. Then, we propose an online stochastic randomized rounding algorithm (ORA) and an efficient online greedy algorithm (OGA) to solve the formulated problem. Extensive simulation results demonstrate the effectiveness of the proposed ORA and OGA algorithms in terms of the minimum application completion ratio.
Scopus Subject Areas
- Control and Systems Engineering
- Electrical and Electronic Engineering
- Service model
- Multi-access Edge Computing