In the news

Oct 29, 2015

Cellular Simplifies its IoT Road Map

By Rick Merritt

LTE Category 0 gets nulled

SAN JOSE, Calif. – The Category 0 option for 4G LTE cellular has fallen to the pruner’s shears. It is the latest alternative to be weeded out of the road map for low power cellular that had become overcrowded in the race to network the world for the Internet of Things.

In the past year a growing set of low power wide area (LPWA) networks have emerged targeting IoT. Sigfox, LoRa, Weightless,Ingenu and others aim to undercut the cost and power consumption of today’s cellular.

In the rush to respond, Altair and Sequans are already shipping LTE Category 1 chips that ratchet data rates back to 10 Mbits/second (downlink) and 5 Mbits/s (uplink). Qualcomm officially announced Monday it will deliver its own Cat 1 chips early next year.

Vendors had planned a follow up LTE Category 0 that would lower maximum data rates to sub-Megabit rates with chips that could have shipped before 2017. However, vendors decided that specification would not offer enough differentiation before its follow-on emerged, Category M also known as eMTC.

“CAT-0 is in fact being skipped, and we have argued it should be for quite a while,” said Eran Eshed, a co-founder and vice president of marketing and business development at Altair. “It offers marginal cost savings over Cat 1, no power advantages over Cat 1 and requires infrastructure upgrades that Cat 1 doesn’t,” he said.

“Furthermore Cat M is now just around the corner, so the motivation for carriers to go to Cat 0 is practically not there, yet we do support it for customers who insist on Cat 0,” he added.

Category M by all accounts is the big win. It will target a max data rate of 100-200 Kbits/s with features that make it more spectrum efficient than what was planned for Cat 0, said Aapo Markkanen, an analyst at Machina Research. Others suggest Cat M could deliver data rates of 400-700 Kbits/s.

Just when Cat M arrives is unclear. Qualcomm would only say its chips for Cat M and a separate class of 3G-based networks called narrowband IoT (NB-IoT) are “expected to align with the 3rd Generation Partnership Project (3GPP) Rel.13,” suggesting it’s unclear when the specs will be complete enough to start building chips.

“I’m being told that the specification for LTE-M (Cat M) is pretty complete and ready to be included in LTE Release 13, but I’ve found it impossible to get a coherent response from the vendors about what it actually will entail,” said Markkanen.

NB-IoT was itself the result of vendors with competing proposals agreeing to collaborate. Markkanen believes it will target max data rates below 50 Kbpit/s and as such be the closest rival to the non-cellular LPWA IoT networks such as Sigfox. In terms of costs, Machina analyst Andy Castonguay believes Cat 1 devices are hitting $18-25, while Cat-M aims for $7-10 and NB- IoT targets $5-7.

“NB-IoT should offer a slimmer [chip] and lower cost than Cat M where use cases require slower and lower power connectivity. We will support both on our next generation chip,” said Eshed of Altair.

NB-IoT is “now a work item that’s meant to reconcile the underlying two main tech proposals, Narrowband LTE and Cellular IoT,” Markkanen said. “My understanding is that as a standard it’s still a delightful mess. The official line is that it’s nearly a done deal, but in reality there’s still a lot of ground to cover,” he said.

Indeed, when asked about the targets for NB-IoT, Intel, one of the companies working on it, declined to comment.

Eshed said NB-IoT is “also known as ‘clean slate’ in 3GPP terminology, meaning that it was architected almost from scratch,” said Eshed, suggesting it is based neither on LTE or 3G.

This week’s news was testament to the race to deliver wide-area IoT nets.

A day after Qualcomm announced its chips, Ericsson said it will demo at an industry event a system using the so-called Power Saving Mode of Cat 1 with help from AT&T and chips from Altair. When asked, an Ericsson representative clarified that “the Power Saving Mode features are planned in the Ericsson SGSN-MME system in 1H 2016.”

Within hours of the Ericsson news, Sigfox announced it is working with the City of San Francisco to install its network there. Aiming to be smarter than other smart cities (a whole other horse race), San Francisco will give Sigfox access to public spots for its base stations and even help cite the antennas.

San Francisco will be “the first of ten U.S. cities in which Sigfox will deploy its network by Q1 2016,” said the company’s release. That schedule would represent a delay of several months in its plans announced in May “to set up 1,300 base stations to cover 10 U.S. cities by the end of the year and as many as 4,000 base stations covering 30 cities by the end of 2016.”

Sigfox is still on track to cover 10 U.S. cities by the end of the year, said Luke D’Arcy, senior director of global semiconductor and module partners for Sigfox. It already has 30 base stations up and running in San Francisco, enough to cover the city. In addition, the company now aims to cover 50 U.S. cities by the end of 2016, he added.

San Francisco Mayor Ed Lee spoke of the potential impact in the Sigfox press release. “Creating a network of this kind, the city will be able to attract new startup companies, strengthen existing businesses and provide more jobs, economic growth and continuing prosperity for our residents,” he said.

It would be hard to imagine San Francisco attracting more Web and IoT startups than it already holds, driving up rent for those living there, but that’s another story. It would be equally hard to imagine any mayor not expressing the same hopes as Mayor Lee.

Click here to read the original article.