Whatever Happened to Cat.1 bis?

Category One Cable, generally referred to as Cat.1, has been the standout success story for cellular connectivity on the Internet of Things (IoT) during recent years. It has been popular with device manufacturers that want global coverage at the most cost-effective price point. While the dedicated cellular IoT technologies of LTE-M and Narrowband-IoT (NB-IoT) have been mainly national rollouts, with LTE-M focused on mobile and NB-IoT on stationary IoT device deployments. Roaming for both has been a secondary concern, especially for NB-IoT where entirely new reconciliation agreements need to be established between carriers based on the number of roaming devices and not the amount of data transported. Conversely, Cat.1 was already available on LTE networks worldwide, ready to be utilized properly for the first time, as its 10 Mbps throughout rate not being useful for consumer mobile broadband but providing plenty of bandwidth for IoT devices.

Cat.1 bis is an IoT-optimized version of Cat.1, its specific innovation being that it mandates the use of a single receive antenna. Previously, the 3GPP explicitly required LTE devices to use two antennas. This reduces device complexity, footprint, and the bill of materials (BoM) costs, further improving affordability. ABI Research’s analysis of cellular IoT module market shipments reveals that since 2016, Cat.1 shipments increased from 0.7 million to just under 43 million in 2020, to constitute 22% of all LTE module shipments and 14% of overall module shipments. Only 2G, >Cat.1, and NB-IoT had higher market shares in 2020, but 2G is now on the decline, and NB-IoT is almost exclusively the preserve of the Chinese domestic market. Cat.1 has a strong future and risks being one of primary inhibitors to 5G adoption in the IoT; yet to date, Cat.1 bis, which seems like a no-brainer to support, has somehow stalled.

Cat.1 bis was specified in Release 14 in 2018, but there has been surprisingly little response from the IoT industry. The first Cat.1 bis chipset was the UNISOC IVY8910DM, which was announced as sampling in mid-2019, and was commercially available worldwide in Q1 2020. The first module to use the UNISOC chipset was Fibocom’s L610, launched in February 2020, with the European band variant being certified by the Global Certification Forum (GCF) as recently as the 17th of December 2020. UNISOC remains the only dedicated Cat.1 bis chipset supplier, but Neoway, MeiG Smart, and Telit have since joined Fibocom in offering modules. Fibocom’s L610 was certified by Deutsche Telekom, Vodafone, and Telefónica in July 2021, but not yet by any US carriers. Western module vendor enthusiasm for the standard is generally negative, with Telit’s LE910S1 being a single antenna version of its existing LE910C.

Cat.1 bis is presently seen as the result of the realization that China’s IoT market needs an alternative to LTE-M, which was deliberately shunned in favor of NB-IoT. But it has now become clear that NB-IoT is not suitable for all low-end IoT connections, especially mobile ones This is the reason why Chinese vendors have been the first to bring Cat.1 bis chipsets and modules to market. But it is interesting that others have yet to follow, with neither HiSilicon, MediaTek, nor Qualcomm seeming to be interested in the standard. Fibocom is notable as being the most international of the Chinese module vendors, with 68% of its turnover in 2020 being from overseas markets, so it could be the most natural party to introduce Cat.1 bis worldwide. But questions over the international acceptance of Chinese chipsets could impede the take up of its current model.

Some module vendors think that outside of China, Cat.1 bis will only be relevant in Europe, the Middle East, and Africa (EMEA) and Latin America (LATAM), i.e., countries where there are also no LTE-M networks. Some also think that the price gap is too great between Cat.1 and Cat.1 bis, with Cat.1 bis modules estimated to be only half as expensive as Cat.1. This puts Cat.1 bis into the realm of low-power wide-area (LPWA), where margins are just too low and entering the market can seem like participating in an inevitable race to the bottom for pricing. Instead, it is likely that NB-IoT-type business models will be needed for Cat.1 bis to be worthwhile to module vendors, i.e., where there is a service-based usage component. Module vendors that have a strong global Cat.1 business do not wish to destroy that market by launching Cat.1 bis. Similarly, vendors whose business is targeted at higher margin products, in the higher-priced markets of Europe and the US, are not showing any interest in releasing Cat.1 bis modules, especially if there are no non-Chinese manufacturers offering chipsets.

As Cat.1 bis module certification is so recent, real world take-up has yet to register in ABI Research’s shipment data collection, and presently Fibocom is the only vendor that would show up. ABI also pays close attention to module vendor product launches with its new IoT Cellular Modules market data. In the first edition, more than 1,000 module models were catalogued as being commercially available at some point during the previous two years, and as soon as new Cat.1 bis models are launched they will be added. Updates are published every six months. Cat.1 bis has the potential to be as widely applied across IoT use cases as Cat.1, but interestingly Fibocom’s own expectations initially position it as being most immediately useful in wearables. Providing a similar footprint to LTE-M and NB-IoT, but with broadband throughput and global coverage for centralized manufacturing and localized activation. Any shipments that occur in 2021 are expected to all be within China.

Western semiconductor manufacturers, principally Qualcomm, are unlikely to make Cat.1 bis chipsets as they do not wish to adversely affect their existing Cat.1 sales revenue. Nor would they wish to impede the already slow take up of LTE-M, which they have made substantial investments in. Even Qualcomm’s NB-IoT-only chipset was slow to come to market and was developed in collaboration with a third-party, as Qualcomm had yet to make adequate returns on its dual-function LTE-M and NB-IoT chipsets with 2G fallback. Similarly, western semiconductor manufacturers do not wish to force the hand of their module vendor customers and take away their own Cat.1 module market. For as soon as the likes of Qualcomm release a Cat.1 bis chipset, western module vendors will be compelled to compete to become the first to release modules, to minimize losses, and secure whatever Cat.1-related busines remains.

Nominally Cat.1 bis seems like a great innovation, but it is not necessarily welcome in a market where hardware is already highly commoditized and where vendors are struggling to translate bigger annual shipments into greater year-on-year revenue. Vendors are already years’ deep into the development of regularly recurring service revenue-based models that incorporate MVNO connectivity and value-added service-based platform subscriptions. Many are also increasingly seeking to only develop module hardware that have strong unit profit margins and refuse to enter into pricing wars. In light of which, Cat.1 bis would seem to make most sense as an addition to most module vendors’ portfolios only once service-based revenues come to constitute the majority of their turnover—and once the inclusion of yet more hardware that is commoditized by design is no longer a problem.