Modeling and Performance Evaluation of Map Layer Loading in Mobile Edge Computing Paradigm

Abstract

Mobile Edge Computing (MEC) is a promising technology to prepare cloud services in the vicinity of mobile radio network. Mobile users could offload their tasks to the MEC servers to acquire results in minimum latency. In this paper, we analyze the performance of map layer loading in the MEC paradigm. We show that how workload, connection failure and service rate could influence on the mean response time and job rejection probability. We extract the service architecture of map layer loading in the MEC platform. Each phase is mapped into an M/M/1/C or M/M/K/C queue. The cyclic inter-dependencies among sub-models are resolved by fixed-point iteration technique. Discrete Event Simulation (DES) is conducted to find numerical results for each sub-model. Finally, the behavior of mean response time and job rejection probability as two performance metrics is studied.