Context-Aware Multi-QoS Prediction for Services in Mobile Edge Computing

Abstract

Mobile edge computing (MEC) allows the use of services with low latency, location awareness and mobility support to overcome the disadvantages of cloud computing, and has gained a considerable momentum recently. However, Quality of Services (QoS) of MEC services are changing frequently, resulting in failures in QoS-aware service applications such as composition and recommendation. Therefore, it becomes critical to develop novel techniques that can accurately predict the QoS of MEC services to avoid such failures. In this paper, we leverage the QoS attributes and three important contextual factors to perform the prediction, as they are highly influential to the QoS of MEC services. Specifically, we propose a context-aware multi-QoS prediction method for services in MEC. We first propose an improved artificial bee colony algorithm (ABC) to optimize the support vector machine (SVM), then we apply the optimized support vector machine to predict the workload of MEC services. Finally, according to the predicted workload and other task-related contextual factors, we predict the multi-QoS of services based on the improved Case-Based Reasoning (CBR). Extensive experiments are conducted to show the effectiveness of our proposed approach.