IEEE Consumer Electronics Magazine - November/December 2022 - 12

Security
Figure 1. Risk assessment on O-RAN threats.
The SFG identifies 52 threats as high risk, nine as
medium risk, and seven as low risk in terms of
impact and likelihood level. Considering the majority
number of threats, i.e., % 76%, in a high-risk area,
a security-by-design approach in O-RAN is of prime
importance.The keypoint is that the threats against
AI/ML are in the high-risk area and need to be a priority
for academics and industries for further investigation
and can be used as a starting point for
determining future riskmitigationmeasures.
AI Threats Against Intelligent Ran
Intelligent radio will use cutting-edge AI and ML
techniques in 6G architecture to address advanced
network deployment. Consequently, for secure
radios, it is necessary to anticipate suspicious
activities carried out by malicious nodes while the
communication processes are being carried out.2
The O-RAN components can collect the acquired
application information and radio environment
information and execute ML models based on the
received information. However, the learning phase
of a ML system can be influenced by poisoning
attacks, which can cause themodel to learn inaccurately.
To give just a few examples, some types of
poisoning attacks include data injection, data
manipulation, and logic corruption. During the
inference phase, evasion attacks employ adversarial
examples that have beenmeticulously crafted in
order to circumvent the model deployed in the ORAN
orchestration module. Attacks on API-based
ML models include model extraction, model inversion,
andmembership inference.
TOWARD SECURE 6G RAN
It is most likely that the 6G RAN will be based
on the O-RAN architecture.3 However, O-RAN is
susceptible to additional security risks beyond
the 3GPP RAN. First, the open architecture can
broaden the surface of threats on RAN. Second,
O-RAN enables radio functionalities to be
dynamically disaggregated. In this architecture,
users are operating in an untrusted environment
in which different network functions use shared
third-party infrastructure to exchange information
between layers. The SFG recommended the
intended countermeasures to tackle the security
issues in 6G RAN related to O-RAN architecture.
The O-RAN provides an opportunity to protect
the network at the edge, thus minimizing the exposure
of network resources to attacks. For fast and
light detection processes, the near-real-time security
countermeasures can take place in an O-RAN
radio intelligent controller located close to the enduser
and implemented as xApp. Moreover, a nonreal-time
attackdetectionmodule canprogrammed
as rApp and deployed to the orchestration module
for more advanced heavy processing on ML algorithms.
These functions enable mobile operators to
run their security applications in real time and use
advanced schemes to monitor and analyze threats
continuously. Security solutions will no longer be
confined to one company's products, making the
telecommunications industry morecompetitive.
Conclusion
Open interface and AI are vital enablers for
the next generation of 6G mobile networks, and
ensuring security is critical for bringing the 6G
vision to fruition. Security solutions that are
both intelligent and robust can be provided by
AI-enabled 6G security. This article identifies
future research directions by discussing challenges
in open and AI-enabled 6G networks.
&REFERENCES
1. Security Focus Group (SFG), " O-RAN security threat
modeling and remediation analysis, " Tech.
Specification, V03.00, 2022.
2. Y. Siriwardhana et al., " AI and 6G security:
Opportunities and challenges, " in Proc. Joint Eur.
Conf. Netw. Commun. 6G Summit, 2021, pp. 616-621,
doi: 10.1109/EuCNC/6GSummit51104.2021.9482503.
3. A. Banchs et al., " Network intelligence in 6G:
Challenges and opportunities, " in 16th ACMMobiArch
Workshop, 2021, pp. 7-12, doi: 10.1145/
3477091.3482761.
12
IEEE Consumer Electronics Magazine
http://dx.doi.org/10.1109/EuCNC/6GSummit51104.2021.9482503 http://dx.doi.org/10.1145/3477091.3482761 http://dx.doi.org/10.1145/3477091.3482761
IEEE Consumer Electronics Magazine - November/December 2022

Table of Contents for the Digital Edition of IEEE Consumer Electronics Magazine - November/December 2022
