The Internet of Things depends on networks that are optimized for machines. Mobile telecom operators' networks have been designed to handle the sort of communication that people have with one another and with various types of content (video, web pages, occasional exchange of fitness tracking data and so on). So it shouldn't be surprising that for many of the applications where devices such as sensors and actuators communicate with control systems and databases, different networks have been used. But as IoT takes off, there is a scramble to provide the connectivity that is needed, and mobile network operators can smell a new opportunity.
Such operators have played in this space for years, and many wide-area communications applications rely on SMS or GSM data services in industries such as water utilities and logistics for monitoring or tracking. But IoT applications are much broader than this, and the economics of using a SIM-based module and a typical mobile data plan just don't add up for many IoT use cases. So new networks built on unlicensed spectrum, and architected to deal with large numbers of connections exchanging small amounts of data, have emerged - notably SigFox, LoRaWAN and RPMA. There are many others, but they are more proprietary and small scale.
Responding to the possibility that these new networks will "eat their lunch" the licensed-spectrum mobile network operators have worked very hard to accelerate the processes of R&D and standardization for a suite of technologies that can compete. Between them, the 3GPP standards-based technologies EC-GSM-IoT, LTE Cat-M and NB-IoT now offer telecom network operators a chance to see off unlicensed competitors by providing the sort of service quality guarantees that are easier to manage if you have exclusive use of spectrum bands, and speed of deployment that is hugely helped by having existing network infrastructure already in place.
Yet technology and coverage are only two of three things that IoT users are looking for: a sensible commercial model (including sensible pricing) is the third - and that's where network operators will need to get it right: traditional ways of pricing data network services won't necessarily work with new IoT applications. Furthermore, there will be some big IoT users - municipal authorities and utilities for instance - whose instinct is to own infrastructure rather than buy network services: their needs must be addressed.
There are signs that operators understand these issues. For instance, AT&T’s offer of 1GB of connectivity (and 500 SMS messages) across up to a thousand devices for $25 indicates a break from a previous fee-per-device model (which suits few IoT applications).
Our report for Heavy Reading (Low-Power, Wide-Area Networking: Opportunities & Options) looks at LPWA candidate technologies for public networks, compares their capabilities and reviews market traction. It also profiles 13 organizations that are providing (or developing) LPWA solutions for customers.