Page 18 - ITU Journal Future and evolving technologies – Volume 2 (2021), Issue 2

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

In [9], the uplink data rate ( ) for computation of‑ Layers of the MCC Architecture: MCC architecture in‑

loading of mobile device user is de ined for the cludes ive main layers: the application layer, the per‑
computation of loading decision ∈ {0,1}, where ception layer, the network infrastructure layer, the Inter‑

= 1 if user chooses to of load the computation to net communication layer, and the computation layer [13].

the cloud, otherwise = 0 if user decides to com‑ The application layer correlates different mobile applica‑

puter its task locally on the mobile device. Hence, the tions; it demands high computational power and is re‑
uplink data rate ( ) is sponsible for delivering end user resource‑demanding

services. The perception layer handles the physical con‑
,

( ) = (1 + + ∑ , ), nection with mobile devices; it relies on the network in‑
2

∈ \{ }∶ =1
(1) frastructure layer to establish a smooth connection to ac‑
cess more computation and cloud applications services.
where is the channel bandwidth, is the user’s The network infrastructure layer corresponds to the layer

transmission channel, , is the channel gain be‑ that handles the con iguration of the physical mobile net‑
tween the mobile device user and the base station work. Besides, it serves as a connection gateway from
, and is the background interference power. the perception layer to the computation layer and rep‑

Besides, the maximum rate of channel capacity in resents the cloudlet infrastructure, which is used as an
an Additive White Gaussian noise (AWGN) channel is edge’s link between UE and the cloud environment. The
de ined in [10] as: Internet layer coordinates the interconnectivity and com‑
munication of the mobile devices and the Internet; it plays

= (1 + ), (2) the role of the link using Transmission Control Protocol
2
(TCP), User Datagram Protocol (UDP) and Internet Proto‑
0
where , and arethechannel bandwidth, trans‑ col (IP) suite to connect mobile devices to the cloud en‑
0
mit power, and power spectral density of the noise, vironment. The computation layer is associated with the
respectively. Thus, providing enough bandwidth for computation phase for of loaded mobile tasks ‑ it includes
data‑intensive tasks is vital in order to minimize en‑ massive storage resources, servers and task of loading
ergy consumption and latency in MCC networks. managers and is in charge of decision making and data
analysis/other real‑time services provided to the UE. An
• Data Analysis: Data gathered from UEs can be pro‑ illustration of the ive‑layers architecture for MCC is pre‑
cessed and analyzed at the edge level to extract es‑ sented in Fig. 2.
sential information. This reduces the latencyof send‑
ing and receiving data to the cloud for analysis.
• Security: Edge computing enhances cloud environ‑
ment’s security at the edge of the networks through
micro‑service management, hardware‑assisted,
caching systems, Software De ined Networking
(SDN) and the use of machine‑learning‑based tech‑
niques. Several techniques have been proposed to
protect vulnerable systems against various attacks
such as Distributed Denial of Service attacks (DDoS),
wireless jamming, spoo ing and man‑in‑the‑middle
attacks.
Security solutions that apply reinforcement learn‑
ing techniques to provide secure of loading to the
edge nodes were proposed by [11]. A deep learning‑
based physical layer authentication that uses spa‑
tial heterogeneity of wireless channels was proposed
by [12], and their techniques distinguish multi‑users
such as legitimate edge nodes from attackers and
malicious nodes without a test threshold. Fig. 2 – The layered architecture of mobile cloud computing.

In order to improve performance of mobile cloud com‑ 2.2 Performance metrics of MCC

puting, edge computing can be enabled in 5G networks
through SDN, network function virtualization, massive Several parameters should be considered when evaluat‑
MIMO, dynamic radio technologies access, D2D Com‑ ing the performance of MCC. Performance metrics are

munication, etc. However, the resources and services determined based on cloud service providers and end

provided by the edge cloud are limited and can only users’ needs. The end user needs seamless network

support a inite number of devices. connectivity, reliable and uninterrupted service provi‑

4 © International Telecommunication Union, 2021