Mighty IoT stand-off looms in 3GPP this week September 14 2015
Rethink Wireless is a sample of our full paid service Wireless Watch, click link below for 4 week FREE trial. Wireless Watch Trial
As growth potential for mobile services shifts from consumers to 'things', it is increasingly urgent for the 3GPP community to control the networks on which that internet of things will run. LTE in its current form is not fit for purpose, so vendors and operators have been working on ultra-low power variants which could support very long battery life and very cheap chipsets, among other key IoT requirements. However, a major split is looming as Ericsson, Nokia and Intel range themselves against Huawei Qualcomm and a group of major operators.
This week sees a 3GPP meeting in Phoenix, Arizona, where many groups and companies will present submissions for future LTE releases and for the 5G program, which the standards body kicks off at the event. On the IoT front, a stand-off is brewing, which represents most of the great rivalries of the industry - Ericsson vs Huawei; Intel vs Qualcomm. And perhaps more telling, the two camps epitomize, on one hand, the traditional state of affairs in cellular - the old 2G/3G vendors leading the way, with a few US operators in tow - or on the other, the new order, with far heavier input from global operators, and a Chinese giant at the helm.
Huawei has pushed the need for a new air interface for the IoT in Release 13, to get around the compromises inherent in repurposing LTE. Its Cellular IoT 'clean slate' proposal claims to achieve a $10 module cost and 10-year battery life. It has harnessed expertise gained with the acquisition of IoT chipset specialist Neul, and gathered considerable early operator support, including that of Vodafone and China Unicom.
Early versions of the platform have been tested by China Mobile for smart parking, and in future, Huawei sees this '4.5G' technology, as it calls it, converging with LTE and moving forward into 5G.
By contrast, Ericsson's Narrowband LTE (NB-LTE) proposal makes far greater use of existing LTE technologies. This is the main conceptual difference between the two approaches, which the backers say will accelerate adoption by leveraging existing technology investments and ecosystems, and so create superior economies of scale to Huawei's approach, which requires new chipsets and may only work fully with the LTE platform in future releases.
The NB-LTE approach is not the surprise - the timing is. Discussion of a way to reuse LTE standards for the IoT have been rumbling for years. Nokia, Qualcomm and Ericsson have all discussed variations on this theme in GERAN meetings and hustled to get support for their particular approaches.
At the start of this year it seemed that none had got significant backing, and last month the two network vendors joined forces and signed up Intel to support their NB-LTE proposal. Sprint and Verizon Wireless were also part of that initiative. That threw a cat among the pigeons, especially for Huawei, which had been taking the high ground with the progress on Cellular IoT.
Ericsson, it seemed, had mainly been concentrating on its enhanced GPRS proposal for Release 13, so the unleashing of a concerted effort to seize the LTE platform too, at this week's Arizona summit, has come as a surprise. The Swedish market leader has succeeded in getting most of the large vendors - Nokia, Alcatel-Lucent, Samsung, ZTE and Cisco - on its side, along with Intel (though fewer operators than Huawei).
Members of the Huawei camp mutter darkly that the Swedish firm's agenda is to create sufficient confusion to get eGPRS adopted as the main solution for Release 13, and push the question of clean slate vs adapted LTE into Release 14, cancelling out Huawei's early mover advantage.
The end result is perhaps more critical for Intel than anyone else, since it represents its chance, at last, to secure a leadership role in a mobile platform, something which has eluded it. In the early showdowns over 4G standards, it sought to push WiMAX, an alternative to LTE whose ecosystem it could control, but backed the wrong horse. That left it well behind in 4G, but it sees the IoT as its chance to get back into the game on equal terms with Qualcomm.
Intel said it will provide a roadmap for commercial NB-LTE chipsets and product upgrades beginning in 2016. Nokia and Ericsson pledged to provide the required network upgrades to support an extension of existing LTE networks with NB-LTE optimized for low power M2M communication.
"We are excited to collaborate with Ericsson and Nokia on the next wave of wireless innovation to connect the growing IoT market segment, and to further grow the momentum for Intel's LTE portfolio and roadmap with NB-LTE," said Stefan Wolff, general manager of Intel's multi-comms business unit, in a statement.
There will be other proposals on the table but they will be overshadowed by these two. There has already been significant work already on making LTE suitable for the IoT - breathing new life into neglected specs like Category 1, moving towards Category 0 and LTE-MTC - there remains little consensus about the M2M-oriented technologies for the upcoming Releases 13 and 14. Some specialists in low power wide area (LPWA) networks, such as Sigfox and Semtech, are presenting their own systems with a view to converging them with the 3GPP program.
This debate is relevant to 5G too. The Arizona meeting marks the start of the 3GPP's work on those standards, and enhancements made in Release 13 and 14 will point the way to future platforms. For instance, the decision to adopt either a clean slate or an LTE-based path will indicate current thinking in the cellular community and may help shape the 5G conversation about continuity versus clean break - though of course, that one will go well beyond the 3GPP's remit and take in IT, virtualization, licence-exempt spectrum and many other topics.