As part of our testing series on our blog (here & here), Marc and I made some tests with the AllWize K1 to see how deep the radio penetration could go. For this, the ideal building was right in front of us: Las Arenas in Barcelona. This historical building used to be a bullfighting arena which has been converted into a huge shopping center, imagine a big cylindrical building made out of traditional bricks covered with some modern building materials inside. What a perfect building to test the radio penetration of our first IoT development tool.
Let me remind you that the main perk of AllWize products is the deep radio penetration. AllWize uses a low frequency, the 169 MHz, which has the power to go through walls. Unlike humans, electromagnetic waves can go through walls, but not without a loss. Though, not all radio waves are the same, some can travel long distances and some can go through thick walls. This is where Wize shines.
This article contains lots of theory, if you’re a techie, go ahead and read it. Otherwise, if you don’t want to go back to your high school physic lessons, you will probably skip the part above the table.
Theory first
The Achieving Optimum Radio Range application report by Texas Instruments 2015, includes lots of very interesting data about the impact of antenna height, the signal path loss or the impact of different construction materials for radio frequencies. In the construction materials table 169 MHz is not covered but you can still get an insight of the loss due to different thicknesses of different materials and how it changes with wavelength.
Frequency absorption is different depending on the material and its density. Some frequencies can even be trapped into obstacles, but this is a whole different story. Energy loss of radio signals going through an obstacle is what we call attenuation and we measure it in dB, that is: the ratio between the incoming and the outcoming energy.
As you can see attenuation increases with frequency, i.e. decreases with wavelength. Thus, we can expect 169 MHz to suffer way less attenuation than 868 MHz (used by Lora and Sigfox). But it also depends heavily on the material. The more dense the more attenuation. Concrete is a radio-killer.
This data is mostly experimental. Physical models are really hard to do but it’s somewhat common sense. Attenuation will depend on two factors: the size of the obstacle and its material. The size will affect frequencies in a different way. The material will affect all of them more or less the same way.
Materials vs frequency
I once read a metaphor to explain why the size of objects impacts radio signals differently. Let’s picture a boy and a mosquito in a rainy day. The boy will be able to run under the rain. Rain drops will barely be a problem for him, apart from the fact that he will end up wet. For the mosquito it will be impossible. Mosquitoes do not fly under the rain since the very first drops will knock them down. The rain drop size is big compared to the mosquito, but insignificant for the boy.
The same happens to different wavelength radio signals - knowing that the lower is the frequency, the bigger is the wavelength. A 20 cm brick wall will be a bigger obstacle for a shorter wavelength than for a longer one. The wavelength for 868 MHz is 34 cm (used by Lora and Sigfox), in the same size range than the brick wall. For 169 MHz the wavelength is 178 cm (used by Wize, hence AllWize), 9 times the thickness of the wall. Here the 868 MHz signal is the mosquito and the 169 MHz one is the boy running under the rain.
Remember my post about the distance tests and the discussion about the link budget which is the balance between the receiver sensitivity and the different gains and losses of the radio transmission. If we have a link budget of around 120 dBm it will quickly disappear with 120 cm of concrete or 60 cm of reinforced concrete. A few concrete walls and you are done.
But walls are not made out of concrete, right? They are usually made of bricks, plasterboard, steel, reinforced concrete pillars or beams. But floors are often done with concrete in modern buildings. This is where we face our biggest challenge.
Walking with antennas
As the people visiting the mall for their early Christmas shopping walked around trying to find that special present, Marc and I were again walking around with antennas poking out of our pockets. He went onto the roof, while I waited having a coffee in the -1 floor. The historical building of Las Arenas is now a big mall of 6 floors and 4 floors of underground car park.
Our setup was very much the same as the one we used for the distance tests. An AllWize K1 as receiver with a Taoglass FW.80.SMA.M Meteor antenna connected to a computer logging all messages, and a second AllWize K1 with a smaller 2J0819-C708N antenna powered by a power bank as transmitter. The overall antenna gain (should I say loss) is around -7dBi.
As deep as it can get!
I didn’t miss a single packet as Marc went up with the elevator (glass walls) to the roof, but once there not all locations were good. The outside roof has a round walk, and the floor must have been thick concrete since no messages got from his transmitter to my receiver.
Finally, we found a good location closer to the center of the building, with open spaces and less structures. It was then my turn to go down to the underground parking floors. Signal reception from floors -2 and -3 was quite good but close to the limit with RSSI values of -108 dBm. In floor -3, we noticed reception depended on the orientation of the antenna, probably a polarity issue mixed with several bounces in the signal. Reception from floor -4 was very irregular and inconsistent. We concluded it was out of the limits for a good radio link.
The results were quite expected, we can go through 8 storeys and several floors of concrete. Next time we will use the AllWize K2 with a more adapted antenna and maybe some other boards to compare different technologies.
The quest for the deepest radio is not yet finished, to be continued…
Interested in our product? Get in touch with us!