Page 17 - ITU Journal Future and evolving technologies – Volume 2 (2021), Issue 2

ITU Journal on Future and Evolving Technologies, Volume 2 (2021), Issue 2

Fig. 1 – Overview of cloud communications: (a) A traditional cloud computing architecture provides access to resources between remote servers in the
locally central cloud and the end users. (b) A hierarchical 5G‑enabled MEC architecture provides access to resources to the end users from remote serves
both in the locally central cloud and more than one mobile‑edge clouds.

delivery for real‑time and computation‑intensive applica‑ UE to the edge cloud; MEC integrates computing sys‑
tions with Virtual Reality (VR) and Ultra‑High De inition tems that provide on‑site computation and informa‑
(UHD) video features. tion processing, which help to reduce latency and
achieve real‑time responses from the cloud. Com‑
Mobile Edge Computing: Mobile Edge Computing putation of loading provides computation solutions
(MEC), also referred to as multi‑access edge computing, to data intensive applications that require high com‑
is a standard that de ines a network architecture in which putational processes. The following parameters are
cloud computing capabilities and services are enabled at considered when performing tasks or data of load‑
the edge of the mobile network. When cloud services are ing: the transmission status between the UE and its
provided closer to the mobile UE, latency and network edge server and the current edge server load status
congestion are reduced and the applications running on [7].
UE perform better. The design of mobile edge computa‑ Tasks in MCC eligible for of loading can be classi‑
tion networks is conceptualized by taking into considera‑ ied into two categories: computation‑intensive and
tion the aftermath of both communication and computa‑ data‑intensive. Computation‑intensive tasks are the
tion. Edge cloud servers are implemented directly at base type of tasks that need heavy computations with rel‑
stations using a generic computing platform for allowing atively fewer amounts of data transfers. The of load‑
the execution of applications closer to the end‑user equip‑ ing decision of these tasks depends on the amount
ment; they act as cache servers as well as transcoding of required computations. Data‑intensive tasks are
servers with a given storage capacity and computing abil‑ the type of tasks that need a large amount of data
ities [5]. The more detailed important roles of MEC archi‑ transfers. These of loading tasks to the MCC envi‑
tecture in 5G networking systems are described in [6] and ronment are vital for the performance of the applica‑
its main services can be summarized as shown below.
tions, and the of loading decision of these tasks heav‑
• Storage: Since the storage capacity of UE is limited, ily depends on the network bandwidth, since a net‑
the edge cloud handles a large amount of delay sen‑ work with lower bandwidth will increase latency and
sitive data generated by UEs in a real‑time manner waste the UE energy [8].
as accessing cloud computing systems directly in‑
creases latency. This can be observed in the following mathemati‑
cal relation, by considering a wireless access base‑
• Computation Of loading: Computational tasks and station , which can be either a Wi‑Fi access point,
processes requested by UEs are of loaded from the femtocell, or macrocell in cellular networks.

© International Telecommunication Union, 2021 3