17 December, 2018 - Those of us with a few years behind us have been talking about smart buildings and home automation for a long time.
Over the last few decades, advances in electronics, control systems, communications and IT have made it possible to improve performance, reduce operating costs and increase the comfort of our homes/offices.
We have seen, for example, the revolution in structured cabling, ICTs, control systems, smart remotes, the emergence of WiFis and now IoT in buildings.
In every era new technologies have enabled the emergence of new communications services and have generally enabled the reduction of operational costs, the improvement of building operations and the improvement of the quality of user services.IoT will be no exception. The availability of low-cost smart devices and large traffic capacities at almost no cost will enable a degree of automation, control and quality of life that we have never known before.
The engineering processes of these IoT projects usually have their own challenges and trade-offs (performance vs. cost). Many of these challenges are already on the way to being solved (bandwidths, miniaturisation of sensors and radios, etc.). However, there are others that, in our opinion, have not yet been adequately addressed.
This is why in this scenario, we develop in this article several lines of work in which our company is specialising and which we consider to be of great interest for the future success of IoT deployment projects in buildings:
1. Cybersecurity
2. QoS in wireless services
3. Location of people and things
Cybersecurity
We believe that the problem of cyber-threats may be one of the Achilles' heels of the technologies behind the new IoT solutions.
Although we have been living with them for some time now on our conventional broadband networks, we cannot say that we have the situation under control, as new attack strategies emerge every day that try to take advantage of this connected world.
Now, with the emergence of the IoT, it will be very easy to deploy multiple smart devices (sensors, actuators, readers, etc.) that will communicate with each other very easily. These new ecosystems are based on interoperability and what on the one hand is an advantage, may in the long run also be a disadvantage. Thus, these open systems could be exploited by hackers to build methods to attack these networks and thus compromise their functioning.
Our vision is that IoT technology will become a consumer product and therefore we are heading towards a scenario of heterogeneous technologies and products, similar to what we now experience with smartphones or PCs.
We can think of many examples. Electronic locks will be bought not so much because they are secure (this will be taken for granted), but because they can be programmed with multiple functionalities or because of their compatibility with our smartphones. Lights will become communication, decorative and entertainment elements. The temperature in offices/rooms will be regulated automatically and according to the needs of each person. Devices will be developed in homes to make life easier for the elderly and/or dependent people.
The main thing is that there will be multiple devices around our environment that will be communicating with systems in homes, buildings, companies and ultimately with people themselves.
This highly connected scenario makes it very necessary to ensure the inviolability of communications and authentication of data.
If IoT solutions are not built with cybersecurity in mind, the same (or similar) threats that we find in our existing IP networks will extend to this world and in this case may be even more dangerous (e.g. think of the regulation of personal health devices).
To deal with this problem our company is currently promoting the use of UTM7 Firewall Software solutions that can be deployed on ARM, Intel and/or cloud architectures.
Here is a diagram of how different sensor networks would be protected.
In this scheme, some of the nodes also become Firewalls and the rest are connected to them via VPN networks with encrypted traffic.
By implementing this type of solution in all types of devices or networks of IoT devices, they can be isolated from the threats that reach them from the internet and it is also possible to define policies depending on each use case or device to be controlled. A typical use case would be to build a simple VPN network to secure the communications of all industrial devices (robots) in a factory or sensors in a building.
QoS in mobile services
In many cases IoT services in buildings will be supported by different wireless networks. Here is a collection of them: 4G, BLE, WiFi, LoRa, Sigfox, ZigBee, Dash7, RFID, NB IoT, etc.
Each of these technologies (and others to come such as 5G) will be fighting for their slice of the market. Users will often have several alternatives to solve their needs and it is very likely that many of them will coexist in the same building, each specialising in certain functionalities and features.
In this scenario, then, we will depend on the proper functioning of these radio solutions for the correct operation of our IoT networks.
To ensure that the deployed networks work properly, we are proposing to the device manufacturers in our projects to include some kind of quality of service measurement system. This is critical for reliable and secure communications.
However, even for something as common as a WiFi network, QoS measurement is not usually taken into account. So we usually end up with "the WiFi network is malfunctioning".
To start raising awareness in the industry, we have developed measurement systems that allow us to stress these networks and thus analyse behaviour under various traffic situations. Until now it was considered that simply having a good signal level (dBm) was enough to have good communications, however, we have shown that it is also very important to analyse the networks with a real load of connections and associated traffic. Thus, in the attached graph you can see the aggregate traffic of a WiFi Access Point (AP) as a function of the load of connected stations. The green line corresponds to an AP with a good performance. The red and purple lines correspond to APs that, although they have good performance with few clients, quickly reduce their performance as more stations are connected.
The conclusion is that not all devices behave the same when loaded with connections.
This same analysis should be taken into account when choosing our Wireless IoT solutions. One thing is that the standards are met, another is that they are interoperable and another is that the features that are in the catalogues are also met. Sometimes the problem is solved with a change in the configuration of the system, but other times, the product (AP in this case) is deficient. As you know, nobody gives a penny for a penny ...
The reality is that designing for coverage alone has proven to be insufficient to guarantee bandwidth. In many cases it is necessary to perform a field verification to check the SLAs if we want to be sure that we are not being taken for a ride.
To facilitate this type of measures in our company we have developed solutions that allow similar and test scenarios of multiple connections over Wireless networks. In the attached photo you can see a device capable of generating traffic of up to 256 stations/devices.
In this way it is possible to certify the wireless networks once they are deployed and validate, in a real way, that the IoT services are working correctly and securely.
To complete the perspective of this point (Wireless QoS), we would like to comment that in this radio network environment, we must consider that our measurement systems should be able to validate that the radio networks are capable of securing communications against all types of threats.
As a paradigm of the problems faced by these systems, we discuss below the case of "man in the middle", which is no less important for being well known. This attack has been widely used since the beginning of data communications networks. A device is placed in the middle of the communications between the victims' computers, which do not detect anything and think they are talking to each other, but unfortunately all the traffic is being redirected through the attacker's device. The attacker thus has access to all the information circulating between the victims' devices.
As radio networks become more ubiquitous, it is easier for attackers to gain access to victims' devices (you don't even have to connect a cable to the network anymore).
Fortunately, there are known solutions to this problem and they are based on the use of secure wireless networks that carry, among other things, sensors that detect such attacks. However, it is unbelievable that at this stage the vast majority of WiFi installations of SMEs in Spain are not protected against this problem! We can imagine what will happen to IoT networks if they become widespread and are deployed without taking this type of threat into account!
This is why we are proposing to our customers to transfer the best practices of secure Wireless networks to the IoT world.
This means not only appropriate investments, but also a design to achieve certain SLAs and/or QoS (performance and cybersecurity). Subsequently, certification/verification of these commitments is necessary. With all the marketing that goes on in our industry, the only way to ensure that networks are deployed correctly is to measure them.
Locating people and things
As everyone knows, it is not possible to use satellite-based location services (e.g. indoors), yet it is in buildings that we spend most of our lives and where the vast majority of our society's economic activities take place. This has led to a frenetic activity in our industry to search for location solutions that work in areas without GPS coverage (shopping centres, warehouses, subways, airports, etc.).
Thus, today, there are multiple technologies and solutions for the location of people and things. The vast majority are based on the concept of using beacons (RFID, WiFi, BLE and UWB are the best known) as a reference.
There is another line of research (in which Aplicaziones has worked) that consists of the use of inertial navigation through the use of sensors applied to the movement of people. This solution minimises the need for beacons as a reference and, by means of algorithms for the fusion of multiple sources of information, manages to locate people indoors at a minimum cost.
This work has been carried out in our company for more than 5 years. Some of our projects have even been successful in European research programmes.
But now, with the deployment of IoT networks, these strategies will have to evolve, as we will soon have multiple low-cost devices in buildings that can be used to provide location services.
So beacon-based positioning engines, typically with TDOA (Time Difference of Arrival) or AOA (Angle-of-Arrival) algorithms, will have multiple network alternatives for location determination.
So we are in for a revolution in in-building location services. It is time to get down to work so that companies and individuals can take advantage of all this innovation.
The implementation of such positioning solutions will enable the emergence of new systems for managing human resources (e.g. facilitating occupational risk prevention processes) and things (controlling the location of valuable assets).
Finally, and to give a small perspective of the indoor location market, we just want to indicate some data from a last study conducted during the month of November 2018 by an American market research company (Marketinsights):
The indoor location and navigation market will reach USD 28.2 Billion, with an aggregate growth rate of 38.2% over the next few years.
I think we have a good excuse to be optimistic about the future of IoT networks in buildings.