How the Internet of Things will communicate

The Internet of Things (IOT) is currently one of the biggest hypes around. According to Cisco the number of connected things exceeded the number of people on earth in 2008 and Cisco even forecasts that the number of connected things will be as much as 50 Billion. Others forecast a bit lower, Business Insider for instance forecast that there will be 34 billion devices connected to the internet by 2020, up from 10 billion in 2015. On the lower end the forecasts range from 20 to 25 billion devices for the year 2020.

No matter who will be right at the end, there will be a lot more things connected to the net and this puts a few challenges on the communication network, especially for the last mile, which is handling the various end-points, such as sensors, meters, actors, and so forth.

The current worldwide retirement of the 2G networks poses a real thread for the IOT from a device perspective. As it stands today the early adopters of IOT, such as smart meter projects, now face the problem that they need to replace the communication modules and one question is now what the best replacement option is? 

Requirements for IOT projects differ from the typical requirements placed on 4G and future 5G networks. For IOT projects we are looking for low cost per end-point, long battery life, low maintenance requirements and they should be easy to integrate. For campus projects Wi-Fi seems to be a good option, but the challenge here is that access points usually do not serve more then 100-200 concurrent devices. So from a network perspective we are looking for high capacity per access point and at the same time the typical IOT project requires a lot less bandwidth than a mobile phone user.

Very many devices send a huge amount of very small messages is in my opinion a simple paradigm that describes the networking requirements for IOT.

Low-power, wide-area networks (LPWAN) may be the right response to the need for reliable, efficient, and easy to deploy connectivity to help unlock the full potential of the internet of things.

There are a few promising contenders in that field. Narrow Band IOT (NB-IoT) is one of these technologies, which is currently defined by the 3GPP, the same group that already defined the 3G, 4G and other mobile standards. 3GPP has developed this standard to enable a wide range of devices and services to be connected using cellular telecommunications bands. The benefit of this technology is that it will most likely be easy for the incumbent operators who can easily roll out this technology on their existing network infrastructure, a potential downside could be that the IOT devices may only be a second class citizen on the network.

A further contender is SIGFOX in Wikipedia it says SIGFOX is a French company that builds wireless networks to connect low-energy objects such as electricity meters, smartwatches, and washing machines, which need to be continuously on and emitting small amounts of data. They focus on simplicity, upload of data (rather than download), really small message sizes and really low costs.

LoRa, another LPWAN technology, is maybe the best backed technology, with companies like Cisco, IBM being part of the LORA alliance, which was started by Semtech. One of the challenge for LoRa may become the different frequencies it operates in, which are not compatible. In the US for instance the frequency is different from Europe and so forth and many legislations these frequencies are subject to licensing. But licensing is still a challenge for all contenders the moment, especially for public networks.

Lastly I have looked into Ingenu’s RPMA network, which is from a technology standpoint a clear favorite of mine for the projects I am typically involved in.

RPMA uses the popular 2.4 GHz band, which is the only spectrum available as a continuous unlicensed band around the world.

Of course, using the 2.4 GHz band means encountering interference. Bluetooth, Wi-Fi, cordless phones, and even baby monitors utilize the 2.4 GHz band so it is no surprise that the 2.4 GHz band has large amount traffic running through it. RPMA was designed to operate in unlicensed frequency and gives Ingenu flexibility with its network to operate reliably in the presence of interference, the ability to cut through the clutter, while still being a good neighbor.

RPMA's uplink data capacity per tower utilizing multiple RPMA channels is 624 kbit/s and downlink is 156 kbit/s. This data capacity and number of devices supported per tower are more than adequate for the vast majority of IoT applications The full two-way protocol, build-in security and the ability to run standard protocols such as MQTT makes RPMA the compelling LPWAN technology.

However, I personally believe that all technologies will come to play and if the forecasts are halfway correct, the market is big enough to provide a space for all technologies.

What do you think?