Page 33
energy resources make latency, energy efficiency, cost, reliability, and security/privacy more desired features (Sisinni
2018).
In traditional IoT solutions, the large amount of IoT data generated by the IoT devices is uploaded to the cloud via a wide
area network for further analysis and data analytics to provide end-user feedback. IoT consists of a multitude of different
devices, sensors and actuators and may even comprise whole cloud infrastructures.
However, there are many challenges due to exponential increase of the number of devices. As supply constraints ease
and growth accelerates, it is expected that by 2025, there will be approximately 27 billion connected IoT devices in the
world (IoT Analytics 2022). In addition, communication costs, bandwidth needs, and latency constraints make cloud
solutions unsuitable for real-time and time-sensitive applications. With these capabilities - low latency, fast response,
context-aware services, mobility, and privacy preservation, edge computing has emerged as the key support for intelligent
applications and 5G/6G Internet of things (IoT) network. Many Artificial Intelligence (AI) solutions based on machine
learning, deep learning, and swarm intelligence have exhibited the high potential to perform intelligent cognitive sensing,
intelligent network management, big data analytics, and security enhancement for edge-based smart applications
(Bourechack 2023). Therefore, using AI applications closer to the edge is a promising solution for achieving high system
performance and improving quality of service (QoS) and quality of experience (QoE) for delay-sensitive applications.
4. IoT Security Risks and Challenges
In 2014, ABI Research has predicted Internet of Things as cybersecurity’s next frontier and securing the Internet of Things
to be a considerable challenge in the next decade. Security implications are more varied than for traditional IT settings
and new variables come into play, including safety considerations, consumer privacy, and data protection (ABIResearch
2014b).
While the IoT developments point to future opportunities, there are risks that arise when people can remotely control,
locate, and monitor even the most mundane devices and articles. Challenges for the realization of the IoT system include
interoperability due to heterogeneous data and variation in data interpretation, scalability (sustainable existing networks
and integrate new networks), security and privacy issues. With various applications based on use of IoT devices - either
consumer, industrial, manufacturing – an organization may have to deal with the security risks and segregate security
countermeasures based on the types of devices, applications, and criticality of the data. The introduction of IIoT devices
in OT environments may require altering boundaries or exposing more interfaces and services.
Security properties of systems are usually described by security models. Typically, these models describe the entities
governed by a specific security policy and the rules that constitute the policy. However, the continual adding of devices
involving different devices, different sensors, and different physical facility security approaches is increasing security
complexity exponentially. Thus, maintaining a holistic security model able to cope with the dynamic changes of IoT
systems is becoming increasingly difficult. Today, no overall flexible, dynamic IoT security model exists capable of
supporting mission-critical systems while simultaneously enabling the expected rapid advances and disruptors.
Generally, IoT systems have been built connecting existing sensors, devices and infrastructure components, as well as
services. Current IoT platforms contain technology solutions from a wide variety of vendors, each providing
heterogeneous components with individual levels of security. Also, the security measures, if any, within the IoT
components have not been designed to consider the dependencies arising from the IoT’s connectivity capabilities or its
data correlation and information retrieval capability.
Many of the current IoT solutions use existing protocols, standards and concepts not designed for IoT. Many systems are
built on the basis of a lack of vision concerning the real potential of IoT. Another challenge is that protocols engineered
for legacy IT and OT components may not operate as intended in current computing and networking environments and
are vulnerable to manipulation. In addition, these protocols have their own vulnerabilities and when referring to make
everything reachable through an Internet connection, there should be improvements or mitigations to current flaws even in
the IPv6 protocol. Also, preserving privacy in today’s IoT is still an open challenge.
Challenges related to the technology of security of IoT include out-of-date software and hardware, the use of default, and
weak identifications. The prediction and prevention of attacks, the difficulty in finding a device that is affected, and other
security and data protection challenges include policies and procedures. There are some other challenges, particularly
legal and ethical. The legal laws are related to cybersecurity, and ethical laws and regulations are related to privacy,
access, and integrity of information, and the compliance of these laws might be a challenge.