Page 29 - ITU Journal Future and evolving technologies – Volume 2 (2021), Issue 2

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

• A Finite Markov Decision Process was proposed to minimize the communication cost
for delivering mobile data with different delay sensitivities through multiple wireless net‑
works and manage replicas by monitoring data requests.
• A Task tracker approach was proposed that can gather resource consumption and of‑
loaded tasks information and determines which task should be executed on a given ser‑
vice provider.
Technique Machine Learning, cryptography, • Not suitable for large networks
deep learning, Gaussian Binary Re‑ since the algorithm is based on
stricted Boltzmann Machine, etc.
binary classi ication.
• Requires exploring all the possibles
states and pairs of actions.
• Model not evaluated in practical
Data Security & and real time environment.
Privacy • Not suitable for ile sharing systems.
[11, 51, 52, 53, 54] Description • Machine learning algorithms were applied for spammer identi ication in industrial MCC.
• Reinforcement learning techniques, especially Q‑learning, were applied to protect edges
from spoo ing, malware, jamming, and eavesdropping attacks that might occur during data
of loading to edges nodes.
• Deep learning‑based solutions were applied to prevent and detect cyberattacks in MCC
online.
• Encryption and decryption techniques were proposed to achieve data security.
• Match‑then‑decrypt technique was proposed in which a matching phase is added before
the decryption phase, to improve encryption and decryption in security schemes.
• A data encryption solution was proposed between mobile devices and private and public
clouds environments.

is irst encrypted by the mobile device using the private

only if the test passes, which is more suitable for per‑

cloud’s public key, and sent to the private cloud, which

forming matching before decryption. The transformation
decrypts it using its own private key and performs user’s

is applied to obtain a chosen cyphertext (CCA2) secure

authentication. When the authentication is completed,

extension. The whole architecture is embedded in the

four algorithms of the anonymous attribute‑based access the mobile device partially computes the cipher text

control system known as Setup, KeyGen, AnonEncrypt, of the remaining data block by block using the private

AnonDecrypt. The matching phase returns the symbol cloud’s public key and sends it to the private cloud, which

to terminate decryption with overwhelming probability, completes the remainder of the encryption and decrypts

it ends with the initiation of the next decryption phase; it using its private key. A data ingerprint is generated

the decryption phase returns the original message. The for every metadata block, which is sent to the public

solution focuses on decryption because the full decryp‑ cloud from the private cloud; the public cloud decrypts

tion cost linearly increases with the complexity of access the message using its private key and performs data

policies. authentication, if the authentication is successful, the

public cloud sends back an acknowledgment message.

With the same vision of enhancing data security in The proposed scheme enhances the con identiality of the

an MCC environment, Yang et al. [54] proposed an iles and the security of the encryption key. However,

with this method, only the data owner can access the

encryption scheme known as File Remotely keyed En‑

ile which is not suitable for ile sharing systems. Table

cryption and Data Protection (FREDP)that performs

1 gives a qualitative overview of different solutions with
data encryption between mobile devices and private and
public cloud environments. In the proposed scheme, their proposed techniques and their shortcomings.

the computation resources of private clouds are used

5. OPEN CHALLENGES AND ISSUES

to remotely encrypt mobile devices data; however, the

encryption key is not shared with the private cloud envi‑
Most of the challenges existing in the MCC environment

ronment which performs data integrity ication; the

today are associated with reducing latency, increasing

encrypted data is encrypted by block then shared with

bandwidth, achieving uninterrupted communication be‑

the public cloud to store it. To enforce security in high

tween a mobile device and the cloud environment with in‑

ic systems, the mobile devices and private clouds

termittent connectivity, assuring constant network avail‑
are assumed to have shared authentication key pairs and ability and heterogeneity, providing data access ef iciency
public keys of each other. User’s sensitive information

and security and privacy during exchange of data, and

© International Telecommunication Union, 2021 15