A Fistful of Standards: Asking What NR-RedCap Is For and Why It Is Needed

Long-Term Evolution Machine-Type Communication (LTE-M, aka Cat.M) and Narrowband-Internet of Things (NB-IoT) were confirmed as 5G cellular Low-Power Wide-Area (LPWA) standards by the 3rd Generation Partnership Project (3GPP) and the GSMA in mid-2018. LTE-M and NB-IoT were standardized in 3GPP Release 13, before 5G New Radio (5G NR) was standardized in Release 15, but have now been incorporated into 5G. Coming at the tail end of the LTE standardization process, the unrealized potential for LTE-M and NB-IoT needed to be protected, as both technologies are only just gaining meaningful adoption. IoT network technologies must be guaranteed to be supported by carriers for decades to come to secure customer confidence in them, as IoT deployments must be assumed to need to persist without human intervention to upgrade or replace them for the duration of their operational life.

ABI Research determined that 75 million LTE-M and NB-IoT modules were shipped in 2020. The IoT is a market of organic growth, driven by business needs to solve practical problems, and not by mere hype for the latest technology. The IoT has a history of thriving on older, established technologies that are highly commoditized and universally available. And despite the advent of 5G, both LTE-M and NB-IoT remain perfectly fit for purpose and should not be tied to, and thereby hamstrung by, the inevitable wind-down of LTE in the medium-term. Proprietary LPWA competitors LoRaWAN and Sigfox do not face the same generational threat of redundancy that cellular does. Cellular development is constantly spurred on by insatiable consumer mobile broadband consumption. Like lanes on a motorway, when more capacity is provided, it is only a matter of time before it is filled up.

5G does have something to add to cellular LPWA, beyond adopting and fostering LTE-M and NB-IoT as the basis for its massive Machine-Type Communication (mMTC) plans. First is Ultra Reliable Low Latency Communication (URLLC), the one killer feature to potentially actively compel migration from LTE to 5G in advance of the need to do so purely for the sake of longevity. URLLC will help to serve mission-critical Machine-Type Communication, which is sometimes referred to as “cMTC”, and was standardized in 3GPP Release 16. Release 17, however, will herald a new 5G category for the IoT, 5G Reduced Capability (RedCap), also known as NR-Light (aka NR-Lite) to capture the demand for sub-5G and supra-LPWA requirements, respectively. NR-RedCap will be for industrial wireless sensor networks and will also provide an evolutionary path for IoT applications currently served by Cat.1 to Cat.4 LTE connections.

Release 17 development is not scheduled to end until 4Q 2022, so we do not yet know the precise form that 5G NR-RedCap will take. But what we expect is that it will combine the higher data rates and lower latency of 5G, with the coverage and cost advantages pioneered by LTE-M and NB-IoT, at a price point that is similar to low-to-mid-tier LTE terminal hardware today. Initial use cases will be based around NR-RedCap’s suitability for live video for the visual monitoring and control of automated processes, drones, co-operative machinery, and remote vehicle operation, and for the acutely time-sensitive reporting of sensor data. It is to be positioned between LPWA and 5G NR, without the need for the connection density of LPWA, and without the need for the baseband modem complexity and brute force throughput of 5G NR.

“NR-Lite” was introduced as a 3GPP study item in mid-2019, becoming known as “NR-Light” by year-end, and now as “NR-RedCap” ever since April 2020. The first NR-RedCap chipsets are likely to be commercially available in 2023, and the first devices are likely to ship in 2024. NR-RedCap is anticipated to have a receive-transmit bias of perhaps 100 Mbps on the downlink and 5 Mbps on the uplink. LTE is asynchronous, too, although the relative difference between its downlink and uplink speeds is less extreme. This makes NR-RedCap an unusual choice for IoT end points, which have a greater need to guarantee the transmission of data than the receipt of it, and nominally seems low in uplink throughput rate for video applications. NR-RedCap would be useful for rapid firmware updates, however, and does seem suitable for another market: wearables and even media consumption devices like tablets.

The consumer electronics market has struggled with its cellular attach rate, even for relatively popular devices like tablets, and for popular brands like Apple’s iPad. Cellular-enabled models are priced at too much of a premium and are targeted at a niche, enthusiast market, with many purchasers of cellular-enabled products choosing to forgo the feature when they upgrade, simply because it wasn’t needed. It has long been sufficient for consumers to tether other devices to a smartphone, or to use connected features only when in range of a Wi-Fi network. NR-RedCap could provide the low-cost hardware, with strong throughput, that the consumer electronics market needs, if tariff plan pricing can be figured out. Samsung and Apple are bound to be interested in NR-RedCap for smartwatches and wearables, and with the high volume of sales that consumer electronics enjoys, compared to the carefully costed and meticulously planned deployment of enterprise IoT devices, we may see a bifurcation of the standard.

But is this all a case of over-imagining the specificity of use cases we must provision for? Are we boiling the ocean to formulate standards that the industries they are designed to serve will not have the patience to wait for, or even the eventual incentive to adopt? Are we just looking for a way to scrub clean the slate of LTE, to present the myriad forms of 5G as a connectivity cure-all? Is it an irony that we need so many stages of development to present a “unified” approach to the comprehensive, horizontal utility of cellular connectivity? And wasn’t everything so much simpler with 2G and GSM, when we had a single standard and 4 frequency bands to serve the entire world? The biggest impediment to 5G in the IoT is 2G’s true successor, LTE, which just like 2G has accidentally found its’ place in the IoT. LTE Cat.1 has been an unanticipated IoT success, coming years after it was first standardized in 3GPP Release 8.

LTE Cat.1 has been organically commoditized, as a sunk-cost technology that has already served its original purpose and could now have a new life, powering a parallel connectivity market. By contrast, the newly developed IoT-specific standards of LTE-M and NB-IoT stalled in uptake, with Cat.1’s 3GPP Release 13 sister specification Cat.0 failing to gain any support. New public network technologies need to display enough upfront potential for the sale of subscriptions to be economically justifiable for rollout. But this is not how successful public IoT networks normally develop where, paradoxically, viability follows affordability, which follows volume. In the case of 2G and LTE, the IoT benefitted from the mobile phone and smartphone markets creating that original volume. And here lies the potential issue for NR-RedCap—it needs to fall as a natural IoT successor to LTE, rather than something that has simply been contrived to take its place.

LTE is the biggest barrier to 5G adoption both in the consumer market and the IoT. In many cases, LTE is simply already good enough, and 5G needs to deliver more for the same price, or less, to displace it. Everything that has been made possible in the cellular IoT has been thanks to the economies of scale created by the consumer market, from spectrum license awards, to network availability, to the price of baseband modems. To have developed any of these from scratch to serve the IoT would have been prohibitively expensive, and today’s IoT market would not exist. The 3GPP needs a migration path away from LTE and does not want the IoT to inhibit this process, which will happen if 5G cannot be a suitable replacement for low category LTE. Hence, NR-RedCap. As a precedent, just witness the longevity of GSM worldwide, specifically because of its sustained usefulness for the IoT.