Abstract In this paper, we consider a bipartite queueing model for multi-access edge computing (MEC), also known as mobile-edge computing. MEC provides computing capabilities within radio access networks located in close proximity to where data is generated. To abstract the relationship between users and edge servers as a graph-based construction, we consider a large-scale bipartite queueing model of MEC using mean-field theory, which is a powerful approach for handling a large number of interacting entities. The contributions of our research are threefold. First, leveraging the result that the superposition of renewal processes converges to a Poisson process via the Palm–Khintchine theorem, we approximate the total offloading process in a tractable form. Second, we determine the optimal offloading policy for users to edge servers, and optimize resource allocation based on key performance measures, such as mean delay and the average number of jobs in each edge server. Third, we conduct numerical simulations to empirically validate our analytical results and compare the threshold-based policy with the stochastic offloading. In the case of a small offloading delay, when the arrival rate is low, immediately offloading a job to one of the edge servers can reduce the mean delay compared to local processing. In contrast, for a large offloading delay, there exists an appropriate balance between local processing and offloading.
Abe et al. (Mon,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: