498
B. Russell
10.5 A Secure Network
The concept of edge networking is being transformed. Cloud Service Providers
(CSPs) are pushing analytics capabilities directly into the physical device and Fog
computing now provides analytics, storage, protocol abstraction, service discovery
and policy management nearer to the edge. This allows new operating models
that enable automation across product lines and makes the network an integral
component of an IoT system. In this section, we discuss techniques to enhance
the cyber security posture of IoT networks. We discuss recommendations for
implementing IoT networks that:
• Limit a bad actors’ ability to gain unauthorized access to network services.
• Limit the ability of bad actors from being able to move laterally through a
network.
• Keep network services available at all times
• Protect the confidentiality and integrity of data processed and stored within the
network.
Network designers should consider the following techniques for implementing a
secure and available IoT network.
10.5.1 Secure Wireless Sensor Network (WSN) Configuration
Wireless Sensor Networks (WSNs) may be used extensively in AI-driven IoT systems. These networks use RF protocols such as ZigBee or ZWave to communicate
between peers and gateways. Configure WSN gateways in a cluster formation to
support handling of heavy loads and automated fail-over. Configure monitoring to
alert on low batteries within edge machines. In certain instances, it may be prudent
to tune the power levels of these edge devices to avoid signal leakage outside of a
defined geographic boundary. Implement physical security controls to restrict access
to the machines.
10.5.2 Segment the Network
A method for limiting lateral movement across a system is to segment IoT machines
from sensitive enterprise applications on the network. To segment the network,
identify categories of machine types and configure Virtual Local Area Networks
(VLANs) for each category.
B. Russell
10.5 A Secure Network
The concept of edge networking is being transformed. Cloud Service Providers
(CSPs) are pushing analytics capabilities directly into the physical device and Fog
computing now provides analytics, storage, protocol abstraction, service discovery
and policy management nearer to the edge. This allows new operating models
that enable automation across product lines and makes the network an integral
component of an IoT system. In this section, we discuss techniques to enhance
the cyber security posture of IoT networks. We discuss recommendations for
implementing IoT networks that:
• Limit a bad actors’ ability to gain unauthorized access to network services.
• Limit the ability of bad actors from being able to move laterally through a
network.
• Keep network services available at all times
• Protect the confidentiality and integrity of data processed and stored within the
network.
Network designers should consider the following techniques for implementing a
secure and available IoT network.
10.5.1 Secure Wireless Sensor Network (WSN) Configuration
Wireless Sensor Networks (WSNs) may be used extensively in AI-driven IoT systems. These networks use RF protocols such as ZigBee or ZWave to communicate
between peers and gateways. Configure WSN gateways in a cluster formation to
support handling of heavy loads and automated fail-over. Configure monitoring to
alert on low batteries within edge machines. In certain instances, it may be prudent
to tune the power levels of these edge devices to avoid signal leakage outside of a
defined geographic boundary. Implement physical security controls to restrict access
to the machines.
10.5.2 Segment the Network
A method for limiting lateral movement across a system is to segment IoT machines
from sensitive enterprise applications on the network. To segment the network,
identify categories of machine types and configure Virtual Local Area Networks
(VLANs) for each category.