LoRa May Not Be for Long Haul at Orange
On the face of it, Orange has made a pretty strong commitment to LoRa, one of a crop of low-power, wide-area (or LPWA) network technologies designed to support more rudimentary Internet of Things (IoT) services.
In November, the French incumbent revealed it was building a LoRa network in 17 of France's biggest cities and would gradually roll out the network on a nationwide basis thereafter. A few months earlier, its venture capital arm, Orange Digital Ventures, stumped up $3 million of the $25 million in funding then raised by Actility , a French company developing OSS and BSS functionality for LoRa deployments. (See Telcos Invest in IoT Tech Startup.)
Yet Orange (NYSE: FTE) has acknowledged that LoRa is far from ideal. As an "open" technology, it holds strong attractions for the service provider over Sigfox, another LPWA technology that is fully proprietary. But this openness combined with LPWA's reliance on unlicensed spectrum is also problematic, admits Luc Bretones, the executive vice president of Orange's Technocentre-named product and design facilities.
"It means everyone can develop what he wants and you may have congestion on the network," he says during a discussion with Light Reading at Orange's recent Hello conference in Paris. "The quality could be poor if too many players need to use the same frequencies."
Quite simply, LoRa does not match up on service quality to longer established cellular standards based on licensed spectrum. And that means an operator could not use it to provide the kind of service level agreements (SLAs) that are critical for organizations in sectors such as health and finance. "LoRa for now is not comparable with 2G or 4G in terms of quality of service and SLAs," concedes Bretones.
Clearly, LoRa does score highly over 2G and 4G in other respects. Orange reckons LoRa is about 15 times as energy efficient as cellular. The range capabilities of the LPWA technology are also impressive. Small indoor lamps Orange has developed as an additional network relay are able to carry signals over a distance of one kilometer in rural settings, according to Bretones. A new scheme will see up to 5,000 Orange employees install these lamps in their homes as a means of improving network coverage, especially for what Bretones calls "deep indoor scenarios." (See Orange Hails LoRa Breakthrough as Bouygues Ups IoT Game and Orange's Olivier Ondet Aims for IoT Supremacy.)
Ultimately, though, Orange's plan is to take LoRa's strong points and integrate these in a future cellular standard. "We are already working with the 3GPP [a specifications body] so that all of the LoRa enablers and features -- low consumption and cost and wide area coverage -- are integrated natively in 2G and 4G," says Bretones. "By the end of 2017, we will be able to do the same with native operator networks."
Indeed, the 3GPP hopes to address a number of different IoT requirements in its Release 13, which is due to be finalized in the second quarter of this year. According to Steve Bell, a senior analyst at Heavy Reading , the technologies that come under the Release 13 umbrella could include NB-IoT, a new LTE radio platform for the low-end IoT market, and EC-GSM, a GSM enhancement that could similarly support low-bandwidth IoT services. Orange happens to be trialing EC-GSM with Sweden's Ericsson AB (Nasdaq: ERIC), which interestingly uses the "deep indoor" expression to describe potential use cases for the technology. (See 3GPP Makes Progress on Crucial LTE IoT Spec and How IoT Forked the Mobile Roadmap.)
Next page: Leaving LoRa
So it seems LoRa might eventually be entirely replaced by licensed technologies (Bretones does not go this far, preferring to say it will be "embedded" in regular spectrum, but he seems to mean the capabilities of LoRa rather than LoRa itself). The question is how disruptive such a process would be. And why, if Orange intends to largely phase it out, is the operator investing in LoRa at all?
The answer to the second question may be that Orange does not feel it can afford to wait for the availability of licensed alternatives as other players get their IoT acts together. Just last month, network rival Bouygues Telecom , which Orange is hoping to acquire, announced IoT plans that also rely heavily on LoRa technology. "It's a race," says Bretones, recognizing the need to gain experience and develop expertise as quickly as possible. "The number of IoT devices using LoRa is blooming." (See Orange CEO Sees 50:50 Chance of Bouygues Deal.)
The LoRa network, then, gives Orange more opportunities to test Datavenue, a data-management platform it launched in June last year, and explore new ways of selling IoT services to consumers and businesses. It is already collaborating with companies including SEB, a manufacturer of domestic appliances and cookware; Harmonie Mutuelle, a healthcare player; and AXA, an insurance specialist, on IoT services that make use of Datavenue's data-gathering and analytics capabilities.
Moreover, in France and Romania, it now provides a number of connected home services under the Homelive brand, some of which might feasibly sit more comfortably on a LoRa network. One option it is currently considering is the sale of LoRa lamps and tracker devices in a single package for customers living in areas of poor network coverage.
"All the knowledge we have gathered through this experimentation and all the assets we have will be very important," says Bretones, when asked if Orange will be able to reuse LoRa investments as it shifts to technologies based on licensed spectrum.
But it might not be all that straightforward, according to Heavy Reading's Bell. Once capital investments in LoRa have depreciated, Orange might face a situation where the LoRa network is much costlier to operate than licensed alternatives. That could leave Orange in a quandary about what to do next, and force it to make some big decisions now.
"They could sell the network and customers, migrate the customers at their expense or, over the next couple of years, invest in a virtualized core that enables them to efficiently run both networks," says Bell. "The reality is that there is no simple answer, but there will be a lot of learning for both operators and customers in the next two to three years."
— Iain Morris, , News Editor, Light Reading