There’s a fragmented choice of wireless technologies for low power, wide area IoT
Bob Emmerson
The telecom industry is good at generating confusion, says freelance IoT and communications writer, Bob Emmerson. LTE Cat-1 services are being launched in the States but not NB-IoT (LTE Cat NB1 is the official term), which only arrives in 2018.
However, NB-IoT is recognised as the international standard for Internet of Things (IoT) networks and it is being rolled out in Europe. So much for a global standard. Do you wait and miss out?
Maybe it’s time to check out the alternative mature technologies like Nwave, Ingenu, Sigfox, Weightless, and the LoRa Alliance. But — and it’s a very big ‘but’ — you should also look under the NB-IoT hood and evaluate the trade-offs.
I’ve been writing about cellular networks from the very beginning and it’s abundantly clear that LTE (long-term evolution to 4G) is a great technology. A brand-new network core and air interface technologies provide a groundbreaking combination of efficiency and flexibility. The efficient use of spectrum lowers costs and enables a combination of high-speed, low-latency transmission with cost-effective, low bit rate services.
However, since I wrote those words the IoT exploded and it was clear that the legacy low bit rate service of 50Mbps was over-engineered. Data rates measured in a couple of hundreds of kbps are more than adequate for most IoT applications.
Right now, Cat-1 is still the technology of choice for new IoT deployments, but a 10Mbps downlink isn’t needed. Cat-0, an IoT standard from 3GPP was promoted for a while but now it’s dead, as is Cat-M1. That was just part of the cellular IoT alphabet soup.
NB-IoT, when it arrives, will offer low complexity, low power consumption and long range. Key characteristics include 180kHz bandwidth and uplink and downlink data rates of about 200kbps with half-duplex operation. However, it seems that low complexity, which is needed to bring down module / device costs, has been achieved via trade-offs. To put it bluntly, they had to be made: it wasn’t possible to simply scale down the overall performance.
One such trade-off is the fact that by design (found in 3GPP Specification TR45.820) NB-IoT is planned to only acknowledge 50% of messages serviced by the wireless technology. This is due to limited downlink capacity. If they acknowledge more, it will hit their capacity much harder than they’ve designed for. Without 100% acknowledgement the technology isn’t suitable for use in many use cases.
Another trade-off is delays caused by the buffering of messages and data in what is called transmit packet aggregation. The 3GPP capacity specifications rely on bundling periodic reports into a single larger transmission, which cellular technology needs in order to handle communication transactions efficiently.
I’m a writer, not a communications expert, but it is clear that Cat-1 and NB-IoT are a retroactive response to a relatively new market requirement. The guys behind Sigfox, LoRa and other LPWAN solutions saw it coming and designed brand new technologies from the bottom up.
That said, there isn’t going to be a killer technology, one that captures the market. And despite the limitations, NB-IoT is backed by the combined resources of connectivity ecosystems such as those of Nokia, Telit and Vodafone.
The author of this blog is freelance IoT and communications writer and observer, Bob Emmerson
Comment on this article below or via Twitter @IoTGN