Page 19 - ITU Journal Future and evolving technologies – Volume 2 (2021), Issue 2

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

sion while moving around, and faster responses from where is the bid vector given by the gateway, is the

the cloud servers. On the other end, the cloud service total maximum bandwidth in the Cloud Service Provider
provider needs to meet user’s requirements while it pro‑ (CSP), is the reserved bid for the CSP, and the total
vides fast and reliable services and reduces the overall number of gateways.
cost of servers and infrastructures. Therefore, the follow‑
ing metrics are major determinant factors of MCC envi‑ Reliability: It is the probability that a mobile device
ronment performance. will perform as intended, so that its functions are satis‑
factorily executed for a given period of time under speci‑
Latency: Latency is the cause of delays noticed by end ied operating conditions in MCC. Thus, the reliability of a
users. There are several techniques to reduce latency MCC setting is de ined by the equation below:
from the UE point of view, among which we have dis‑
played local animations, background loading to hide la‑
tency or pre‑fetching and parallel connections on multi‑ = ∑ ∗ (1 − ( + − , , )), (6)
ple threads. Latency occurs as well within the MCC en‑ =0
vironment, delays can arise anywhere from the edge to where
the data centers. Besides, there are techniques for reduc‑

ing the average end‑to‑end delay in the MCC environment ( + − , , ) = , (7)
including machine learning and adaptive priorities based + −
on when the request was initiated. The average latency of with
a device to upload its computation task to a base station = ( ) = ! (8)
is de ined by [14] !( − )!
∗ and
, = , ∗ , (3)
,
,
,
= ( + − ) = ( + − )! (9)
where , is the input data‑size (in bits) for processing + − !( + − − )!
the computation task of the − ℎ device, is the length
of one Time‑Division Multiple Access (TDMA) frame, , for > and 0 otherwise. Here, is the number of avail‑
is the expected channel capacity, and , is the time slot able paths, is the number of actually used paths, is
resource for each device. the maximum number of failure paths, and the number
of failed paths [17].
Energy Consumption: Network energy consumption
in UEs is mainly observed during task of loading, task ex‑ Service Availability: It refers to the state of being used
ecution or computation. If an edge cloud, associated with or obtained, such that MCC availability is directly propor‑
the base station to which the UE is connected, executes tional to its number of active edges and BS. It is essential
its UE’s task, then the computation energy consumption for every MCC systems and mandatory for cloud service
is proportional to the changed capacity of the edge cloud. providers. Besides, it is actually one of the key factors
If the central cloud executes the task, then the consumed to procure seamless data exchange in MCC environments,
energy can be de ined by the energy consumption of the thus there can be interruptions of services or low of data
cloud which the edge cloud is associated to[15]. The en‑ if the system is not 100% available.
ergy consumed by a node is de ined by [16]
( ) = ∗ + ∗ , (4) Quality of Service (QoS): It is the measurement of the

response of a system to different requirements, stan‑
where ( ) is the absorbed energy by the node after a dards, and objectives expected by end users. Thus, it de‑
given time, and are the number of transmitted and notes the level of performance, reliability, and availabil‑

received packets, respectively, and are constant fac‑ ity offered by a system. Moreover, QoS is sometimes as‑
tors based on the energy model.
sociated with Quality of Experience (QoE), which is de‑
ined by techniques such as Mean Opinion Score (MOS),
Bandwidth Utilization: Bandwidth is the measure of Net Promoter Score (NPS) or Standard deviation of Opin‑
the capacity of a channel to transfer data in a network. ion Scores(SOS).
The wider or greater the bandwidth, the greater the
amount of data that can be transferred and the number
of users that can be handled by the network. Therefore, Security: Most of mobile devices contain end‑user per‑
it is vital to maintain high bandwidth in order to achieve sonal information such as pictures, a list of contacts, fre‑
seamless communication in MCC networks. The available quent locations, payment information, etc., which are tar‑
bandwidth of a channel is de ined by [10] geted by attackers. Unfortunately, most mobile devices
are unprotected and vulnerable due to their limited re‑
( )
= ( − ), (5) sources in terms of computation and storage, so that they

∑ =1 cannot run powerful protection systems.
( )

© International Telecommunication Union, 2021 5