Recently, a representative[1] from Google reiterated the special characteristics (low bandwidth and inexpensive) of IoT applications and the need for a brand new network. Over the past few years, several initiatives have been launched to address the potentially massive, low data and low power requirement segment of the M2M/IoT market; Neul and SigFox are two examples that spring to mind.
It was therefore interesting to hear of a development that has been working its way through the 3GPP standards process. LTE, traditionally associated with high data-rate mobile services, is now being engineered to address M2M/IoT applications. This is not a case of data-intensive video surveillance or digital display applications but the use of LTE in moderate data rate, low cost and long battery life scenarios.
At its recent Analyst and Developer Conference, Sierra Wireless described its efforts in piloting the standardization of LTE-M (as it is referred to) over the past three years. The company also described certain design goals for LTE-M and how these are being achieved:
Design for low-mobility situations (e.g. in homes or ‘places’) means that less onerous link-budget specifications can be used to deliver throughput rates of up to 1Mbps, significantly improving in-building coverage.
An explicit goal of the standards activity is to reduce the bill-of-material costs. Sierra Wireless claimed that the long run cost trajectory for LTE-M should approach the current cost profile for 2G modems. Interestingly, a report [2] published through Cambridge Wireless echoes this message. The co-authors of this report (A Choice of Future m2m Access Technologies for Mobile Network Operators) includes mobile network operators and technology vendors. It notes that module pricing should target levels of $5/unit and lower in the long run.
By adapting the connectivity duty cycle, the connected device service life should be extended to 5 years from a standard battery power source. Should these objectives be met, the scope to apply LTE-M for IoT-like applications is projected to expand the addressable market quite considerably. Using market analyst projections, the authors of the “Future m2m Access Technologies” report consider that the addressable market could be expanded considerably through substitution and market expansion effects as illustrated below.
If this is indeed the case, LTE-M would appear to have far-reaching implications for connected device OEMs as they decide on their connectivity technology strategy. Should they adopt present-day, low data rate and long service-life solutions? Is there a time-to-market issue that forces a near term adoption of a hybrid solution based on short-range wireless technologies (e.g. Bluetooth, Wi-Fi etc.) allied to wide-area connectivity (fixed, mobile)? What are the commercial implications of anticipating a technology migration when LTE-M becomes available (sometime after 2017 before high-volume economics lead to long-run cost targets being achieved)?
For all its investment in the standardization process, Sierra Wireless also has a challenge. It has to find a way to differentiate itself competitively once the LTE-M market takes off.
Finally, are the advocates for a brand new network for the IoT correct? That’s debatable and it would be prudent not to ignore this piece of history. Barely five years ago, there was quite a marketing push by Clearwire in the USA for a network to handle high data-rate applications for an emerging class of connected devices. Very little of this materialized. The underlying technology did not generate the scale economics or the innovation necessary to deliver significantly low-cost components globally.
[1] Google Exec: Internet of Things Requires ‘Brand New Network’ (June 2014) - http://www.lightreading.com/iot/google-exec-internet-of-things-requires-brand-new-network/d/d-id/709381
[2] A Choice of Future m2m Access Technologies for Mobile Network Operators (March 2014) - http://www.cambridgewireless.co.uk/news/article/default.aspx?objid=46213
13 July 2015 update
ReplyDeleteThe vendor community is investing greater efforts into LTE-M for IoT applications.
Link to SK Telecom and Ericsson cooperation:
http://www.koreaittimes.com/story/51748/sk-telecom-joins-lora-alliance-promote-internet-small-things
In June (2015), SK Telecom entered into a Memorandum of Understanding (MOU) with Ericsson, a global leading provider of telecommunications solutions and services, to collaborate on the development of LTE-based Machine Type Communications (or LTE MTC) technologies and its related business.
Link to Huawei initiative:
http://telecoms.com/431321/huawei-china-unicom-serve-up-cellular-iot-with-smart-parking/
Chinese institutions China Unicom Shanghai, China Unicom Research Institute and kit vendor Huawei have developed a collaborative partnership to develop a pilot of cellular IoT (CIoT).
Based on Huawei’s “4.5G” LTE service for machine to machine communications (LTE-M), the carrier has rolled out a trial mode of what it refers to as “Smart Parking”, and claims it to be the first commercial application of commercial LTE-based cellular IoT.
8 Nov 2016 Update
ReplyDeleteSigfox undercuts AT&T, others with $2 IoT module
Sigfox aims to undercut its rivals in the emerging IoT market with the introduction of “ultra-low cost” modules that start at $2 each.
The French firm said it had simplified the requirements for the new chipsets, enabling it to offer them at “a historical low.” The modules are available in Europe, the Middle East and Africa for $2 each, and in Asia and the Americas for $3 each.
Indeed, Sigfox’s new chipsets appear to be significantly less expensive than some other modules on the market in these early days of the IoT. AT&T recently outlined plans to launch LTE M modules for as little as $5 each, and chips for other Low Power Wide Area (LPWA) networks are often in the range of $5 to $10.
Sigfox boasted that its new modules are as much as 20 times cheaper than LTE-enabled modules and five times less expensive than other competing technologies.
http://www.fiercewireless.com/wireless/sigfox-undercuts-at-t-others-2-iot-module
10 Jan 2024 update
ReplyDeleteBroadband IoT more-massive than ‘massive’ IoT – way less than short-range IoT
The total number of cellular IoT connections will reach around three billion at the end of 2023, reckons Ericsson, with most, and the increasing majority, connected on so-called ‘broadband’ cellular IoT technologies LTE (4G) and 5G. The arrival of reduced-capability (RedCap) 5G through 2024/5 will further boost this ‘broadband’ segment, said the Swedish vendor.
The twin narrow(er)-band IoT technologies NB-IoT and LTE-M (Cat-M) will continue to take share from legacy 2G and 3G, but will remain off the pace set by LTE and 5G – the mainstays also of the mobile industry’s private networking push. The latest instalment of Ericsson’s bi-annual Mobility Report, available here, says 1.6 billion (53 percent) of the total cellular-IoT forecast will be on LTE and 5G by the end of the year.
But the IoT findings make clear something else, as well: that cellular IoT represents, and will continue to represent, a fragment of the wider IoT market. The total number of wide-area IoT connections, which counts non-cellular IoT connections on technologies like LoRaWAN and Sigfox, will reach 3.3 billion by the end of 2023 and 6.6 billion by 2029, growing at a compound rate (CAGR) of 12 percent – which is the same in the period as for the cellular IoT sub-segment.
The non-cellular share of wide-area IoT will go from around nine percent (300 million connections) in 2023 to around 7.5 percent (500 million) in 2029. These other camps will take issue with the figures, clearly. But most significantly, the total IoT market is worth 15.7 billion connections already (2023-end), meaning cellular IoT is only responsible for a fraction of it (about 19 percent).
And the rest of the market – dominated by short- and mid-range technologies like Bluetooth, RFID, and Wi-Fi, notably, plus a number of proprietary 805.14 mesh technologies – will grow at 17 percent in the period to 2029, by comparison, upping its collective share from 79 percent in 2023 to 83 percent in 2029. Perhaps Ericsson’s next Mobility Report will include a RedCap forecast, to see if that makes a difference to the niche tale of massive IoT.
https://www.rcrwireless.com/20231130/internet-of-things-4/broadband-iot-more-massive-than-massive-iot-way-less-than-short-range-iot
Report at https://www.ericsson.com/en/reports-and-papers/mobility-report/reports/november-2023
22 June 2026 update
ReplyDeleteLoRaWAN eyes ‘fourth pillar’ status as IoT apps start to stack
Yegin responds: “We are more confident than ever before that LoRaWAN will be the fourth pillar of the wireless industry. I mean, just look at the numbers; look at all the applications. Outside of China, there are more LoRaWAN connections than anything else –125 million, growing 25 percent per year. And NB-IoT, which China mandated for wide-area IoT in 2019, is losing steam everywhere else. AT&T has pulled the plug; so has Docomo.
“And LTE Cat 1 bis is eating its lunch; the cellular community is cannibalizing its own. And you have an application here, an application there, and really nothing in the way of massive-scale adoption. Whereas LoRaWAN supports everything from rhino tracking to water meters, street lights, facilities management, industrial monitoring. LoRaWAN has the widest set of applications, not only among LPWAN technologies, but across all wireless communications.”
So 125 million, say, is a drop in a much bigger ocean, of well over 30 billion wireless devices, and not much bigger in the IoT field, which harbours about 22 billion of those. These numbers are from different sources, and used only for context; they discount wired IoT devices, also. But niche-upon-niche, short-range IoT, served by Bluetooth and Wi-Fi, commands the waves.
Again, rough maths, but about four in five (78 percent; 17.4 billion) wireless devices – which might fairly be described as IoT units, such as home assistants and fitness trackers – connect on these shorter-range technologies; the rest (4.93 billion), like rhino trackers and water meters, are on longer-range systems, sometimes served by LoRaWAN. But nominally, even this wide-area category might be cut a couple of ways: for throughput and criticality, at least.
And when those twin requirements are limited and pragmatic, then LoRaWAN works well – and works, also, for a wide range of applications. “We have built a technology that serves numerous applications for massive IoT – as opposed to critical IoT,” says Yegin. More categories, more niches – as defined, in fact, in 3GPP circles. He continues: “You need a licensed band for critical IoT to guarantee millisecond latencies for real-time comms.
“Like for piloting a drone or controlling a surgical robot, or for connected cars and planes – where you need to carry megabytes and gigabytes of data. That’s where you need cellular IoT. Everything else – which doesn’t have to be enveloped inside millisecond latency; where kilobytes of data per device per day is good enough – is massive IoT. And that’s basically more than 95 percent of IoT applications. LoRaWAN is serving that market.”
Some salt, again: this is 95 percent of the wide-area IoT market, maybe, where low-power (LPWAN) versions of cellular go up against a rogue’s gallery of alternative IoT technologies, mostly using unlicensed spectrum bands. LoRaWAN just looks like the best of these LPWAN contenders; its fourth-pillar claim/ambition, and its spiralling run rate, might be taken in the context of all of this. “Until LoRaWAN, those were the three pillars,” says Yegin.
He goes on: “You might say 125 million is niche – compared to all the billions of Wi-Fi and Bluetooth and cellular devices. But momentum shows this is not a single niche. A niche is what you get with Z-Wave, right? The smart home, and that’s it. Or with Wi-SUN, which is two niches: street lights and electric meters. Our numbers are great, especially that growth rate (25 percent per annum); the variety of applications we’re supporting is more striking.”
https://www.rcrwireless.com/20260622/internet-of-things/lorawan-fourth-pillar