IoT Network Visibility | Pipeline Magazine

By: Miguel Carames

With the IoT population expected to outnumber non-IoT devices by the end of the decade, it’s time for communication service providers to change the way they think about their business and network. IoT requires greater precision and network intelligence that recognizes that IoT will create a hyper-segmented ecosystem. To fully capitalize on the opportunity behind connected things, all aspects of the technology stack must come together in synergy, and to do that requires sharp network visibility that enables communication service providers to solve difficult problems. Questions like: Do you have the network intelligence to detect low-power IoT, how do you identify mission-critical IoT versus consumer IoT, or what is your level of confidence that mobile IoT is performing as required? The era of IoT is here: Do you have the necessary network visibility?

We are in a phase of massive IoT growth. The numbers speak for themselves:

• 5.8 billion: Number of licensed mobile IoT devices by 2030, up from 3.5 billion in 2023 (GSMA).
• 46 PB (petabytes): Amount of global IoT data generated in 2028, up from 21 PB in 2024 (Juniper Research).
• By 2028, wholesale IoT roaming will reach $8 billion, implying a CAGR of 16% from 2023 to 2028 (Kaleido Intelligence).

In this context, it is important for CSPs to ensure that all corners of their networks are equipped not only to meet IoT requirements but also to help identify opportunities to maximize revenue. These are the most important questions CSPs need to ask.

Kaleido Intelligence predicts that wholesale IoT roaming will reach $8 billion by 2028. However, the variety of devices and use cases makes it difficult for CSPs to fully monetize this opportunity. For example, energy-efficient IoT devices and sensors, commonly used for remote monitoring of environments such as smart cities, agriculture, and manufacturing, are increasingly joining the IoT roaming category. Juniper Research predicts that there will be more than 490 million energy-efficient IoT roaming connections by 2028, a 560 percent increase from 2023. However, because these devices consume less data and have intermittent connectivity, they can be difficult to identify, which is especially true for devices that are coming into a visited network. This results in a huge potential revenue loss for carriers, especially when you consider Kaleido’s prediction that by 2028, IoT roaming connections consuming less than 10 MB of data per month will account for 59 percent of the total IoT roaming connection base.

Figure 1. Roaming control

A few years ago, GSMA defined the Machine-to-Machine (M2M) Roaming Transparency Initiative. However, practical experience proves that detecting low-power devices is possible only with AI-based deep analytics. AI-based analytics is required to detect, monitor and analyze billions of IoT devices connected to the network to create a 360-degree view of the IoT ecosystem. This in turn provides visibility to CSPs to monitor the performance of the network and services delivered to IoT devices so that they can provide better connectivity and experience to every IoT device.

But it’s not just device detection that’s important. CSPs also need to understand the different device categories to ensure that the right traffic and pricing policies are consistently applied. This is where IoT Packet Printing plays a key role. It combines carrier-grade security, behavior analysis, and traffic analytics to automatically identify the device category, and can also detect misuse and prevent fraud or abuse in the event of IoT device tampering or failure. Traffic and pricing policies for IoT are often more complex than those for retail.

Therefore, communication service providers need to ask themselves the following questions: Does our policy management module have the AI ​​and automation capabilities to manage the complexity and scale of the Internet of Things?

Roaming control is another area where IoT exhibits greater complexity than traditional telecommunications services (see Figure 1 above).