Mobile Edge Computing

Abstract

46 1. Introduction During the last decade, growing and expanding mobile devices, such as smartphones, have accelerated evolution of Information and Communication Technology, especially cloud computing and wireless communications. In cloud computing, currently, many cloud services, such as file sharing, Content Delivery Network (CDN) service, and machine learning, are provided by several cloud vendors, such as Google 1) , Amazon 2) and Microsoft 3). In wireless communications, the mobile devices equip various network interfaces, such as Long Term Evolution (LTE), Wi-Fi and Bluetooth, and can connect to "the cloud" anytime and anywhere. As a result, the mobile devices, especially smartphones, are selected as a main platform of mobile services, and nowadays, the smartphones' owners can easily experience rich multimedia applications, such as 4K video streaming, Augmented Reality (AR) and Virtual Reality (VR), via the wireless networks. Moreover, not only the smartphones, but also sensor devices and surveillance video cameras are connected via the wireless networks. In Internet of Things (IoT) services, the sensor devices, including network cameras, are installed everywhere, and enable to monitor city streets, social infrastructures and nature environments in real time. Because the sensor devices have less computational resources and battery, the sensing data and captured videos are uploaded and analyzed in the "centralized" cloud 40). This behavior is a typical example of "computational offloading." The computational offloading indicates that computational tasks are transferred to external computing environments, such as cloud servers, via wired/wireless networks and executed in the external computing environments instead of an own device. According to the Cisco's report 4) , Cisco forecasts that the mobile traffic will increase seven-fold from 2016 to 2021, and mobile video accounts for approximately 80% of all mobile data in 2021. Thus, due to the centralized cloud, the plenty sensor data and rich contents may trigger the overloaded computing and severe network congestion, especially in the cloud-side backbone network, and invoke longer latency for data exchange between the cloud and end devices. To reduce the computational load in the cloud and to reduce backbone network traffic, cloudlet based computational offloading is proposed 5). Cloudlet (like a mini-cloud or a private cloud) is deployed in the physical proximity to users, such as in a shop and a restaurant, and accessed by using Wi-Fi (i.e., Cloudlet covers a small area). Although the cloudlet may provide the low-latency network connection, this tiny cloud has few computation and covers only few users. Recently, to cover a larger region, and to provide low-latency connectivity and resourceful computing, Mobile Abstract Recently, to provide a low-latency mobile computing platform, Mobile Edge Computing (MEC) is proposed. In this paper, we first summarized the feature capabilities of MEC, such as content distribution and caching, computational offloading and...

Proactive content caching by exploiting transfer learning for mobile edge computing

“…Thus intuitively, we can design new intelligent content caching and distribution mechanism by leveraging the data storage and compute capability of MEC. In particular, MEC servers, which operate as local content delivery nodes and serve cached contents, may perform in a cooperative way and effectively share their caching information with each other . Naturally, the cooperation among multiple distributed caches may enhance the overall network performance significantly.…”

A Data Stream Processing Optimisation Framework for Edge Computing Applications

“…In addition, Beck et al, propose a taxonomy of Mobile Edge Computing (MEC) [3]. They classify potential mobile edge computing applications based on 4 metrics of feasibility: power consumption, latency, bandwidth utilization, and scalability.…”

“…The toolkit was implemented on top of the OMNeT++ framework 2 [22] employing the native network simulation capabilities of the INET framework 3 . OMNeT++ enabled the development of a modular, component based toolkit to simulate the inherent characteristics of edgecloud streaming applications and networks.…”

Toward Anycasting-Driven Reservation System for Electric Vehicle Battery Switch Service

“…CSs are set up as Mobile Edge Computing (MEC) [20] servers with information mining, aggregation and sharing of EVs' reservations with each other. Such a feature is deemed as a scalable solution to the long-term introduction of EVs, in terms of communication cost and system scalability.…”