Navigating the Non-cellular Sea: LPWAN and what to do about it
The IoT is a tricky thing to master. First things first, you have to figure out what it is you want to measure, and choose (or develop) the right sensor for the job. Then you have to understand the data that they produce, perhaps using a platform to smooth things over, and create plans from this data (and maybe even use a separate device to help you plan). Once these things are decided, you can scale up by linking these sensors together and creating a network of data transfer, an ecosystem if you will, that you also need to understand and most likely link with the cloud for even more data analysis. Phew.
Different Ports in the Storm
But really, the most important part is what joins all these ‘things’ together. Besides the technology needed to monitor or track your ‘thing’, what really runs the IoT is the connectivity, the medium by which all this data flows and floats around. Right now there are a variety of options, which cater for different needs. Bluetooth and WiFi are useful for super-short range applications, like NFC posters that you scan with your phone to register your interest, but have a long battery life because they only send tiny data packets and need to stay in that poster for a long time. Cellular connectivity, the one your smartphone uses to go on Facebook, comes in different bands, 2G, 3G, 4G and soon 5G, which have faster speeds, bigger data packets, and higher power consumption the further up the spectrum you go. The big advantage of cellular connectivity for IoT is that it is available around the world, and the spectrum it uses is licensed, which means that the signals will not interfere as they pass each other.
Hoist the LPWAN Sail
These different types of connectivity clearly have different uses, but the truth is that the majority of IoT M2M devices only need small amounts of data to be sent at irregular intervals, over a fairly wide range, and with a long battery life so they can stay in the field generating data. The problem with this is that the cellular band that most devices rely on, 2G, has been decommissioned by AT&T in the US, and other networks are expected to follow, to free up bandwidth for the super fast 4G that mobiles and tablets crave. Enter LPWAN. Low-Power Wide-Area Networks (LPWANs) operate on lower frequencies than cellular, which means the data packets are smaller, they can travel further, and the devices that use this technology can last for years or decades on an AA battery. 85 of these networks were founded in 2016, and it is fair to say that the LPWAN renaissance is well under way. These bands are unlicensed so anyone can use them, which means that developing an LPWAN technology is not as difficult as say creating a new version of cellular, but there is also a fair amount of interference which the main technologies have overcome to differing degrees with different technological quirks.
The Low-Flying Frenchmen
Sigfox, the first of these non-cellular networks to emerge in 2010, uses very low data transfer rates, between 18-36 bytes per second, and as a result has very limited ‘downlink’ communication, when the central data store replies to the device that sent the data to it. Because it uses such low frequencies, Sigfox effectively operates under the radar of other interfering traffic, avoiding signals getting crossed by doing precisely that: avoiding them.
Pieces of Dat-a
Ingenu’s RPMA technology uses a different method, stacking its signals across all the bandwidth allocated to it, avoiding traffic by sending lots of tiny bits of packets over lots of different channels, effectively breezing past ‘bigger’ signals that use one channel per packet. This method also gives Ingenu a huge range, as they can operate on a higher frequency spectrum which is more widely available, and uses a higher energy signal which travels further. LoRa is an alliance formed to make LPWAN more available and accessible to any network developer, as long as they use LoRa hardware. They use a different but similar approach to RPMA called a chirp spread spectrum, which spreads the signal over the entire allocated bandwidth but also increases the frequency of the ‘chirps’ as they travel, meaning that they do not interfere with other signals with fixed frequencies, and do not lose signal strength as they travel.
Standard for the Seven Seas
Weightless is another alliance whose goal is to bring LPWAN to the masses, but they have an entirely open source structure, and have several different versions operating on different frequencies to expand the range of applications they can cover. However, like all LPWANs, Weightless would prefer people just accept their standard, as they are not interoperable with other, more established, networks, and as a result their vision of one global standard is just as likely as any of the others gaining dominance, ie. not likely at all.
To confuse things further, there are also cellular versions of this wide-area technology, NB-IoT (Narrowband IoT) and LTE-M (the M stands for machine), which use the licensed spectrum to almost completely eliminate interference, a spreading technique to avoid ‘channel noise’ that slows down the signals, or operate at the extremes of the LTE spectrum to keep out of the way. NB-IoT has just been standardized by the GSMA, but developers need to fully upgrade their devices to adopt it, which kind of defeats the purpose of using licensed spectrum that is available wherever you need it. LTE-M will apparently solve this issue, but is nowhere near ready yet, and as with all these cellular equivalents will require a lot more power than LPWANs do.
So what will the future hold for connectivity, if not a Super Smash Brawl-type tournament to decide a victor? The answer is not far off. LPWANs will continue to develop independently to different incompatible standards, and will be used much as they are today by different applications that need slightly different things. The cellular alternatives will almost certainly take off in a big way, but again will not satisfy the needs of long-term devices that cannot be removed to change a battery every six months. The advent of 5G will shake up cellular bands, and many applications will move to wide-area offerings to combat sunsetting of 3G, and maybe even 4G bands.
Release the Connectivity Kraken!
So the answer is finally clear: all you need is a connectivity that is currently widely available, flexible enough to compete with new technologies in future, and won’t need to be replaced as soon as a better alternative comes along. Well, luckily, such a solution exists, in the form of multi-IMSI SIM cards that can access several IMSIs (International Mobile Subscriber Identities, or network profiles) at once to remain in the field even when bands are sunset, and switch between operators to get the best possible signal wherever you are. Until and even past the rise of the LPWAN, this flexible cellular technology is the best option for keeping your devices afloat and generating revenue for as long as possible, without being tied to one connectivity type, forced to watch as the IoT tide surges past.