A practical guide to cellular IoT technologies – part 1
This is the first installment in a 2-series of articles on cellular IoT technologies such as LTE-M and NB-IoT. This series is not so much intended for radio network specialists, but rather aims to provide practical, actionable information to users of the technology, focusing on use cases, basic operating principles and performance parameters.
There will be 2 installments in this series:
- Instalment 1 (“What is cellular IoT?”) will focus on the key characteristics of the two main cellular IoT technologies – LTE-M and NB-IoT – their performance characteristics, use cases and comparison against other connectivity solutions
- Instalment 2 (“What are the performance characteristics of NB-IoT and LTE-M?”) will quantify the performance of each technology in terms of signal strength, latency and signal attenuation.
But let’s get started with our first instalment.
What is cellular IoT?
Before we start, a note of warning: the telecoms industry operates with a veritable alphabet-soup of acronyms. These can intimidate the uninitiated. If in this blog, we refer to these acronyms, we will aim to unpack them and explain what they mean in basic, human language. The focus is on being pragmatic and useful for people who need practical rules of thumb. The goal is not to provide a technically exhaustive or complete guide for radio engineers.
What is IoT and how does it differ from normal connectivity?
Let’s start from the beginning. What is IoT? IoT is the acronym for “Internet of Things” – it describes the idea of connecting things (as opposed to people) to a network so they can send (and receive) data about their operation. Typical IoT devices and applications include water meters, doorbells, asset tracking, predictive maintenance, parking sensors, etc….Typically, any device or service that has the word “smart” in front of it will be an IoT device. It is clear that this kind of application will be heavily reliant on wireless connectivity. But IoT wireless connectivity has different requirements from the traditional, “human-centric” connectivity. The table below lists the most important ones:
| Requirement | Importance for human connectivity | Importance for IoT |
| Power consumption | Medium People care about their battery life, but phones can get recharged quickly and without much incremental effort (e.g. at home at night) | High Operators of IoT devices such as water meters can ill afford to go out regularly to replace batteries. Batteries need to last multiple years |
| Bandwidth | High the ability to see the latest tik-tok video in HD is dependent on large quantities of bandwidth | Low Measuring devices do not send cat videos. The data is typical a set of parameters on the functioning of a device, which in terms of payload is actually quite small |
| Latency | High No one likes a laggy gaming or video call experience. Anything more than 100 ms spoils the experience | Low While for some devices, the information transmission needs to be real-time, most devices and services can tolerate a lag of a few hundred milliseconds or even a few seconds |
| Coverage | High No one likes to be out of coverage | High No IoT operator likes to jump in a van to install a measuring device and find out it is not working due to lack of coverage |
Table 1:”human” vs “machine” connectivity requirements – Source: Teragence
In essence IoT connectivity is less stringent in its demands in terms of bandwidth and latency, but equally or more demanding when it comes to power consumption and coverage.
How do cellular IoT technologies optimise for coverage and power consumption?
IoT connectivity piggybacks on some basic physics laws to optimise its performance, specifically Hannon’s law. Shannon’s law in simplified terms states that there are three interlocking parameters which can be manipulated to influence the performance of a radio connection: power, frequency and bandwidth. These relationships can be verbalised as follows
- Higher transmitter power results in more data (i.e. bandwidth) being transmitted or the signal being carried further.
- Lower frequencies (for the same power and bandwidth) carry further.
- A lower bandwidth carriers further than a higher bandwidth (for the same transmitter power and frequency).
As a practical example, 5G standalone will typically operate in the higher frequency bands (3.5 MhZ) where it can push through lots of bandwidth, but in very short ranges (for a given power).
If we look at the other end of the range and want to push through small-ish amounts of data over long distances, then a lower-range spectrum band is ideal. And that is why the typical spectrum band for IoT connectivity is in the lower spectrum bands, i.e. 800 MhZ (Band 20) and 1900 MhZ (Band 3).
In addition to spectrum optimisation , cellular IoT technologies have a few other optimisation tricks up their sleave in terms of the channels within the band used, the sleep/wake cycle of the modems (summarised: instead of being connected all the time, only connect when needed ), etc… All these deliver additional bandwidth and power benefits. Typical battery life for a Cellular IoT device is 15 years. This gets us to another acronym: LPWAN , or “low power wide area networking”, i.e. wireless networks that rely on low-power / low-bandwidth applications. Besides cellular IoT, other LPWAN technologies include Sigfox and LorRaWan. However, both technologies rely on their own proprietary transmitters being installed and therefore lack the ubiquity of cellular IoT technologies (which ride on the back of the cellular connectivity provided by the mobile network operators).
Let’s pause here and summarise: Cellular IoT (Internet of Things) is part of a family of LPWAN (Low Power Wide Area networking) technologies, i.e. wireless connectivity optimised for low power, long range and smaller data payloads suited to connect IoT devices and services. Cellular IoT rides on the back of the connectivity infrastructure provided by the mobile operators and should therefore be available in a reasonably ubiquitous manner, as opposed to other LPWAN technologies such as Sigfox and LoRaWan.
Flavours of cellular IoT: LTE-M and NB-IoT
Having established what cellular IoT is, let’s take a look at the different flavours of this technology. There are currently 2 kinds of cellular IoT technology: LTE-M ( which stands for “Long Term Evolution Machine Type Communication” also sometimes referred to as “CAT-M1”) and NB-IOT (“Narrow Band IoT”). We will not discuss the detailed technical differences between these two technologies, but rather their functional differences, as put forward by the industry and its standards and outlined in the table 2 below.
| LTE-M | NB-IOT | |
| Throughput | High: 300-350 kbps | Low: 30-50 kbps |
| Latency | Low 100- 150 msec | high 200 – 10,000 msec |
| Range & penetration | Medium Upto 11 km outdoor and average indoor penetration (still better than standard cellular) | High Upto 15 km outdoor and good indoor penetration |
| Mobility | Yes – cell handoff | No – no cell handoff |
| Roaming | Available | Not (yet) available |
As a shorthand, NB-IOT is perfect for stationary and indoor applications such as smart metering smart agriculture, parking control sensors, etc…LTE-M is more suitable for things that move around, i.e. medical devices and wearables, asset tracking, POS (point-of-sale) terminals, etc…
This suitability for different applications needs to be considered in conjunction with the availability of the technology in different geographies and with different operators. The GSA (Global Mobile Supplier Association) is typically considered to be the global authority on cellular IoT connectivity availability.
What is 5G Redcap?
The underlying assumptions in the previous paragraphs are that IoT is about different flavours of low-bandwidth, high-latency, long battery life applications. While this is true for many IoT applications, it is not true for all of them: autonomous vehicles, drones and VR applications will need high throughput and low latency and while battery life is a considerations, it is probably not on the timescales of LTE-M and NB-IoT. So 5G Redcap (“5G Reduced Capability”) is a cellular IoT technology being put forward as the “goldilocks” flavour of IoT connectivity, keeping the “just right” balance between “classic”, full-blown cellular 5G on the one hand and cellular IoT connectivity delivered by LTE-M and NB-IoT. We should highlight that 5G Redcap is in the early stages of the product lifecycle and we expect it to take a few years before it becomes as ubiquitous as LTE-M and NB-IoT.
Conclusion
To summarise: machines and sensors’ connectivity requirements differ from “human” connectivity requirements : generally less data, higher latency and longer battery life.
LTE-M and NB-IoT are cellular LPWAN technologies that respond to that need, each with their own specific strengths and weaknesses
In our next instalment we will take some of the claims of each technology and test them to see how they behave in real life.