France, which hosts both low energy wireless network company Sigfox and the LoRa Alliance, is regarded as a pioneer in the Internet of Things (IoT) field. France’s ground-breaking efforts nevertheless run up against US thinking, which is adamant in dismissing these networks as inventions with no future, along the lines of the now-defunct Minitel videotex online service and the mothballed supersonic aircraft Concorde.

IoT networks set to collapse under the weight of connected objects?

The IoT-oriented networks are known as LPWans, which stands for Low-Power Wide-Area Networks. Whether they use technology developed by the LoRa Alliance – an open, non-profit association of companies that are on a mission to standardise LPWans – or Sigfox solutions, these networks, which have been developed specifically to enable connected objects to communicate, are enjoying increasing popularity worldwide. Toulouse, France-based Sigfox claims coverage of 1.6 million km2 across 26 countries, while the list of members of the LoRa Alliance continues to grow. These networks are now booming and analysts predict that Sigfox could very soon join the elite ranks of ‘Unicorn’ companies – i.e. startups valued at over $1 billion.

While few deny their current success, the latest Smart City / Smart Grid fair held at the Paris Expo Porte de Versailles exhibition centre on 8-9 November saw an animated discussion on the topic of IoT networks, and not least the future of LPWans. Among the questions raised during the debate was how they will behave in the coming years, given that the number of connected objects is set to rise exponentially.

The networks and the objects

Rémi Demerlé, Head of Development at analogue and mixed-signal semiconductors supplier Semtech, a company which is developing LoRa technology, reminded the audience that "LoRa technology was initially developed for water meters, which are usually located in basements and for which 2G technology couldn’t provide a satisfactory cost/return ratio. LoRa was developed in order to solve this problem and Veolia [a French company whose activities include water management] has now rolled out the solution on five million meters. Today Veolia sends and receives 130 million messages a day and this massive deployment demonstrates just how useful it is to have a technology that sends information directly."

Another benefit of LoRa technology is the wide range of choice available to users, who can choose at will among manufacturers of compatible modules and network operators, and can even opt to set up their own antennae. An antenna costs only about €500 and you do not need a user licence.  

Nevertheless, Patrizio Piasentin, Head of Southern Europe for Austin, Texas-based Silicon Labs, a firm which claims to develop ‘energy-friendly solutions for a smarter, more connected world’, stressed that too much freedom is dangerous, arguing that you cannot allow any and every manufacturer and operator to just install their antennae as they see fit and deploy millions of connected objects, without being able to guarantee a minimum service quality. "At any given moment, an antenna can only listen to some tens of thousands of messages, with a reach of several dozen kilometres. But what would happen if every French person were able to use only one single connected object? All the tests currently being carried out are in areas where there are very few connected objects but the laws of physics dictate that an antenna cannot receive information from more than 10,000 objects," he pointed out.

Sigfox and LoRa: inequality as regards saturation risks

IoT communications experts stress that it is important first and foremost to understand that Sigfox and LoRa are not equals when it comes to the risk of saturation. Although the two technologies use the same frequencies, Sigfox is an Ultra Narrow Band type of network while LoRa uses Spread Spectrum technology, i.e. the entire spectrum. Cyrille Le Floch, founder of France-headquartered next generation IoT operator Qowisio – a startup that builds low-cost connected objects and also operates its own LPWan on these frequencies – explained why he chose an Ultra Narrow Band network such as Sigfox: “The receiving limit of a LoRa antenna is around 10,000 connected objects, but for Ultra Narrow Band that rises to 100,000.” He reckons that in the future when there are millions of connected objects, and tens of thousands in a given area, this technology will enable them to avoid having their signals jammed.

However Rémi Demerlé pointed to solutions to the problem of congestion on the frequencies used by Sigfox and LoRa. "There are two solutions to the problem of congestion and density of connected objects in a network cell. The first tool the operator can use is deterministic simulation. You can calculate precisely what will be the exact zone of coverage of each of your antennae depending on the geography of the place, the position of each building," he told the Smart City/Smart Grid event audience. This will enable operators to increase the density of their antenna network in areas where they are most needed. This approach could go a long way. US cable operator Comcast is planning to equip its boxes with a LoRa picocell in order to connect all the communicating objects in your home.

Sigfox and LoRa only temporary solutions?

Moreover, there are new technologies on the horizon that will compete, at least to some extent, with the ‘Made in France’ LPWans. One example is the cellular networks that are at last learning to talk to connected objects. 4G is now being developed with the aim of accommodating connected objects using LTE-M (Long-Term Evolution machine-to-machine) and NB-IoT (NarrowBand IoT).

"NB IoT is being sold as a ‘killer app’ for connected objects,” acknowledges Rémi Demerlé, explaining: "It communicates much faster than LPWan technologies and also has advantages in terms of latency. It can therefore meet some particular needs such as those of alarm systems. However, rolling out NB-IoT will require even denser network cells. Today, the LTE network covers only 20% of France. Operators only deploy it in areas where there is a high concentration of smartphone users, not necessarily where there are IoT needs."

Patrizio Piasentin has no doubt at all that part of the IoT market will shift to LTE.  He sees LoRa technology as only a temporary solution being promoted by cellular network operators in order to keep Sigfox in check until LTE’s development for IoT needs becomes fully operational. "Seeing Sigfox appear on their doorstep and without any solution to compete, French operators decided to use LoRa as an interim system. But this is only a temporary solution; as soon as service quality starts to falter, they’ll be able to offer LTE to their customers. We believe that LPWans are temporary solutions, while we wait for LTE for IoT." It is even quite possible that LTE networks will be brought into line with the new standards during 2017, simply by updating the software, he suggested.

4G set to kill off LoRa and Sigfox with NB-IoT and LTE-M?

Cyrille Le Floch argued that there will not only be one sole winner in this game. "We’ll have multiple technologies at our disposal to meet a wide variety of needs. The choice of technology will be closely linked to the Total Cost of Ownership of the connected object, i.e. the cost of the various components right through to the network service, throughout its lifetime," he told the Paris Expo audience. Meanwhile Sigfox recently announced a move to reduce the price of its chips. You can now connect a product to the Sigfox network with a chip costing just two dollars. Little by little, a hierarchy will be established quite naturally, with the cheapest LPWans for connected objects and LTE and the various technologies built on top of them. So where will the dividing line be drawn? Price, energy consumption, plus also the actual quality of service provided by each solution will determine the boundaries for the different manufacturers, Le Floch predicted.

By Alain Clapaud
Independent journalist specialising in the new technologies