From RFID chips to connected objects
The dream behind the internet of things (IoT) was to make everyday objects a part of the digital world. “RFID (radio-frequency identification, up to 10 meters) technology, which is now highly mature and widely used, has established itself for use in connecting large numbers of objects - travel passes, books on loan from libraries, burglar alarms on buildings, that sort of thing”, recalls Nathalie Mitton, head of the FUN project team at the Inria Lille – Nord-Europe center. “If you give an object a unique ID and then read the data collected using a reader, it will automatically be shared online.”
A whole host of connected objects then emerged and were democratised. These objects use proven network technology, such as Wi-Fi networks or Bluetooth, to communicate with modems installed in homes or in offices. Objects in this category range from connected watches and scales to alarms.
Sensors popping up everywhere
What’s new now is that the internet of things has extended to cover the entire physical world, with a whole host of sensors collecting data which is then integrated into digital interfaces. You have sensors collecting data on the environment, for example, including things like temperature or atmospheric pressure, and others collecting data in buildings for the purposes of controlling heating, air conditioning or lighting systems ; explains Nathalie Mitton.
Some of these new objects are within the range of modems, making it easy for them to share the data they collect. But others operate in environments which aren't covered by conventional wireless or wired networks, such as ships’ holds or fields. “One of the aims of our research is to find ways of collecting data in these environments”, explains Mitton, before going into detail on a project linked to connected agriculture her team has been involved in with researchers from Stellenbosch University in South Africa.
Long- and short-range transmission
This is something that LoRa and Sigfox, two network technologies developed in France, can both be used for. These long-range data exchange systems were designed specifically for the IoT and borrow mobile networks (this is also known as cellular IoT) to send small amounts of data at low speeds. Like all modern technology, these systems have their drawbacks, and are only suitable for use in certain situations. “The long range - up to 100 km - has a negative impact on speed, and it is not possible to send data or images in very short time frames”, explains Nathalie Mitton. Another possibility which has been the subject of much research “involves routing data from sensor to sensor until it reaches a point connected to the internet”.
Shared and resource-constrained networks
Since 2013, the 15 members of the FUN project team have devoted their research to self-organising Future Ubiquitous Networks, which is how the team got its name. Their research has mainly been focused on networks used for wireless sensors, the aim being to design network architectures capable of organising themselves and communicating in restricted environments.
What sort of restrictions are they under? Most often, these sensors don’t connect to the major networks of telecoms operators, or to radio frequency bands controlled by certain organisations, such as the fire brigade or the military. Instead, they share small frequency bands which are open to everyone: Wi-Fi, mobile networks, visible light frequency bands, etc. The drawback to these, however, is that they are rare and cluttered. These are ISM bands, for industrial, scientific, and medical use.
Two main obstacles have to be overcome in order to use these bands. You first have to try to limit the risk of delay and electromagnetic pollution caused by interference. One way of doing this is to try to send the most useful data at a time when the network is least busy. You then have to find ways of securing communications on networks that, by their very nature, are totally exposed.
When a malicious node is detected, the network normally has to be able to adapt its routing protocol to allow it to route information via a more secure channel”, outlines Nathalie Mitton.
The FUN is in eliminating constraints
In addition to concerns raised regarding data protection, “sensors are often also small objects, with limited resources in terms of battery life, processing power or memory, and in some cases they are mobile, including those fitted to drones or worn by animals”.
This wide variety of resources and contexts has a number of different implications for our researchers. Faced with limited storage capacities, they must find quick ways of transmitting information so as not to risk losing it, and often have to draw on their creativity. “Our aim is for all transmission technologies, whether light communication, Wi-Fi or LoRa/Sigfox networks, to coexist and even to assist each other.”
As for whether or not such discoveries might be made obsolete by the much-heralded arrival of 5G, which is capable of covering large areas at very high mobile speeds, this specialist is not so sure. There will always be uncovered areas, whether in the ocean or in remote terrain. It’s also likely that there will be more and more sensor networks looking to take a cheaper, alternative route should the opportunity arise.
Two projects for the IoT of the future
|The FUN project team has close ties with this EU project, which is coordinated by the Portuguese consultancy firm PDMFC and partly funded by the Hauts-de-France region. Its aim is to secure communication infrastructure between sensors, which is viewed as critical.||
The researchers from FUN are also involved in another EU project, with support from a number of institutions, including the ANR (the French National Research Agency). Their goal is to invent new techniques for dynamic network resource allocation.
Further reading (only in French)
- Forget the Internet of Things-it's the Internet of Everything, Binaire, 17/04/2019
- A video of Nathalie Mitton presenting the connected agriculture project she's involded in South Africa, Fondation I-SITE ULNE, 08/01/2020.
Podcast (in French)
Sensors for our fields