Abstract
An important use case of the Mobile Edge Computing (MEC) paradigm is task and data offloading. Computational offloading is beneficial for a wide variety of mobile applications on different platforms including autonomous vehicles and smartphones. With the envision deployment of MEC servers along the roads and while mobile nodes are moving and having certain tasks (or data) to be offloaded to edge servers, choosing an appropriate time and an ideally suited MEC server to guarantee the Quality of Service (QoS) is challenging. We tackle the data quality-aware offloading sequential decision making problem by adopting the principles of Optimal Stopping Theory (OST) to minimize the expected processing time. A variety of OST stochastic models and their applications to the offloading decision making problem are investigated and assessed. A performance evaluation is provided using simulation approach and real world data sets together with the assessment of baseline deterministic and stochastic offloading models. The results show that the proposed OST models can significantly minimize the expected processing time for analytics task execution and can be implemented in the mobile nodes efficiently. (C) 2020 Elsevier B.V. All rights reserved.