While preparing for a guest lecture for students who will soon graduate as IoT system designers, I had a chance to reflect on how my own 30+ year journey in wireless technology began.
As a newly baked sales engineer, I was asked to put together a wireless system to replace telephone modems in a SCADA system. The phone lines were connected to each and every pump station in the area, and some 30+ phone lines were costly. During an update of the old modem setup, the local water supply company wanted to look at a more modern solution. Wireless — almost unheard of in the early 1990s.
The first IoT systems before the internet was even a thing
With some help, I put together a draft system design with a plan on how to interconnect all the stations with wireless links up to 10–12 km or more (vague memories here). The antenna analysis and installation instructions were faxed to the customer — there was no public internet at that time. That part I remember clearly. I even had to go to the office on a Saturday just to forward a fax from my antenna supplier to the customer with all the antenna mounting heights and types.
Anyways, we sold the system. The customer got it installed. Everything ran smoothly after some final adjustments.
What did we achieve?
- A system to remotely control pumps for fresh water.
- A system that could read all the substations in a second or two.
- A system where you could restart anything or adjust anything without travelling.
- A system that provided the customer with more insight than before.
The only thing that disqualifies this as an Internet of Things solution was that there was no internet involved. It was like one big isolated intranet with no risk of viruses in the email.
The only anxiety we had during the installation was range anxiety. RF range anxiety, that is. Would the antennas be sufficient or not?
I get it. This is a story from the past. Why bother, right? In the context of the students, I believe in the power of sharing your mistakes so others can learn. We remember the term IoT being coined in the late 1990s, but stories like this are almost untold. The mistake? Around the same time that Kevin Ashton coined the term, we had a Houston moment. A problem.
Range solved one problem — and created another
The payoff time for these wireless links was ridiculously short. About a year. The number of frequencies was limited, but the range was excellent. So with great range, systems that reused the same band in adjacent regions, and a glitch in the radio planning from the Swedish Post and Telecom Authority — tada — we got interference between systems.
System A asked, “How are you doing?”
And system B responded back to system A.
Not exactly a great way to build wireless networks.
So what can we learn from this?
Frequencies are a natural resource. Use them carefully and don’t transmit over a larger area than you actually need to cover. Free frequencies are free for everyone, so there is no guarantee that your message will reach the intended recipient. In these systems, we eventually solved problems through more accurate radio planning, directional antennas, and lower antenna installations. We had almost a full 2 MHz band to play with. Using free frequencies, you will not have that luxury.
The potential problem with free bands is congestion. I tried to raise this question more than a decade ago. In 2012, I published an article arguing that license-free spectrum might become a bottleneck if we continue to rely on it without expanding the available bands. At the time, the discussion was mostly about Wi-Fi congestion in the 2.4 GHz band. The truth is that it can take 10–15 years to free up spectrum globally. And today there are surprisingly few discussions about expanding license-free spectrum, at least as far as I can see.
Success is what crowds a frequency band
Fast forward to 2026. Today the conversation looks slightly different. IoT networks such as LoRaWAN are also built on license-free spectrum. The openness of these bands is what made rapid innovation possible. Anyone can deploy a network, connect sensors, and start building services.
But the physics of radio has not changed. As more deployments appear, more devices transmit, and more gateways are installed, the same question returns: how much traffic can a shared band really carry before interference becomes the limiting factor? This is not a criticism of LoRaWAN. It is simply a reminder of how radio works, and something that I feel is very important to share with the soon-to-be IoT system designers. Shared spectrum, like any shared resource, always comes with shared responsibility.
And if history teaches us anything, it is that success eventually fills the spectrum it depends on. Every successful wireless technology eventually runs into the same limit: physics.