Adaptive Replication for Real-Time Applications based on Mobile Edge Computing

Abstract

With the popularity of mobile devices, data transmission between mobile devices and cloud datacenters not only introduces tremendous data computation and network traffic but also causes high transmission delay. Mobile edge computing (MEC) is proposed to improve Quality of Service (QoS) for mobile applications in 5G networks. However, the data consistency for mobile applications still relies on replication schemes. Frequent updates may thus result in high communication over-head and long transmission latency. In this paper, we design an adaptive replication scheme for real-time mobile applications by considering the number of failed requests and the read/write request ratios. Our scheme adaptively allocates replicas to share loads among MEC servers. It can avoid overloading MEC servers and shorten latency. The simulation results show that our scheme can reduce the response time of replication requests to increase QoS performance.