3D Printed WiFi Reflectors Custom Designed for the Building

Are you a wizard at antenna design? Chances are you’ve never even given it a try, but this tool could change that. Most home-made WiFi signal boosting antenna plans around the Internet share one feature: they are directional antennas or reflectors. But WiPrint is a tool for designing custom WiFi reflectors that map to the specific application.

If we want to increase the signal strength in two or three different locations the traditional solution is an omnidirectional antenna. The problem is, although a good omnidirectional antenna increases the signal power in those locations we want, it also increases the signal power where we don’t want.

A team of researchers led by Dartmouth College created WiPrint to allow users to input a floor plan, the location of the WiFi access point and a desired signal map into the system. The software uses an optimization algorithm to produce a custom reflector shape for that floor plan. The reflector can then be fabricated and placed next to the access point antenna to reflect and concentrate the signal in the specified area, while decreasing signal strength outside of it. The best thing is: you can actually 3D print the reflector and just glue tin foil on it!

The results show that optimized reflectors can weaken or enhance signals in target areas by up to 10 or 6 dB, respectively, and resulting in throughput changes by up to -63.3% or 55.1%. That is not the only advantage, as the researchers point out:

Our approach provides four benefits. First, it provides strong physical security by limiting the physical reach of wireless signals, hence creating a virtual wall for wireless signals. Second, it relies on a low-cost ($35), reproducible 3D reflector, which can be easily replaced upon substantial changes in the environment or coverage requirement. Third, it offers an easily accessible and easy-to-configure solution to non-expert users. Users only need to specify coverage requirements and a coarse environment model, with which our system computes a reflector shape tailored to the built environment. Finally, it is applicable to commodity low-end Wi-Fi APs without directional or multiple antennas.

The sad part is that, for now, no software is available. The study and results have just been presented at ACM’s BuildSys 2017. It would be great to see something like this open-sourced. Meanwhile, this is further proof that [Brian Benchoff] knew what he was doing when he told you to use duct tape for superior WiFi range.

34 thoughts on “3D Printed WiFi Reflectors Custom Designed for the Building

  1. No, that’s not physical security, it’s hoping your signal is weak enough for an attacker not to notice it or be able to receive or connect, that’s security by obscurity. Attackers with highly directional high gain antennas will merely have to get a bit closer or find the strong lobe that lets them be 3x as far away.

    1. I was thinking exactly the same as I was reading it, that their first point is marketing hype to the technically illiterate “it provides strong physical security by limiting the physical reach of wireless signals”. Now if it was one step of an overall series of steps in a system to bring external signal levels below Johnson–Nyquist noise levels for 1 Hz bandwidth at 0.006 kelvin then it would be a different story. So dropping the external signal level below -220.8 dBm then you have actual security.

      1. Realistically, security of a system is made up of a series of mostly imperfect layers.

        Reducing the strength of a wi-fi signal can be one layer of defense. It is real security — not perfect security, and certainly not adequate as the only layer of defense, but it is real security. It really does reduce the chance of an attacker compromising the network, by making it harder to get physically close enough or by requiring attackers to have more specialised equipment.

        It’s not “security by obscurity”. Security by obscurity is security that relies on the attacker not knowing how the system works — e.g. an encryption system that becomes trivial to decrypt when the attacker sees the source code. Physical signal attenuation improves security even if the attacker knows exactly how it works.

        1. Double the diameter of a parabolic reflector antenna get an extra 6dB of gain, or lower the temperature of your first LNA with Liquid nitrogen for almost an extra 6dB. Trying to attenuate a signal is trying to obscuring it.

  2. Very misleading article. Implies ” 3D printing” what is basically an antenna – ergo – needing a metal 3D printer. When in fact, it’s using a ‘regular’ COTS 3D printer to produce what is essentially a platform to apply aluminum foil to.

    No mention of what effect “krinkled” aluminum foil has – vs – perfectly smooth (with no surface imperfections) foil. All the microwave projects I’ve worked on, any sort of blemish would distort the RF field. There is a reason phased array radars have smooth surfaces (one of them being, the math is simplified). Indeed, an interesting adjunct these researchers should explore is perhaps a mini phased array that would permit configurable coverage pattern tuning.

    1. About wrinkled reflectors: I always thought that any imperfections smaller than 1/20 of the wavelength didn’t matter when designing RF reflectors. In the ARRL handbook, this rule of thumb is referenced when describing mesh reflector parabolic reflectors. Likewise in optics, a mirror with imperfections smaller than 1/6 is considered good and 1/20 excellent. So kringled alufoil shouldn’t have any measureable effect on the antenna? From the picture in the article, it appears that the dipole on the router is too close to some part of the foil to reflect any signal properly, so I suspect something else is going on there and the algorithm designing the shape uses some kind of destructive interference to shape the lobes in a very unorthodox way.

  3. Sadly, this software is not likely to ever see the light of day. It is frustrating to see how much of computer science academia seems completely uninterested in actually advancing the state of computer science.

    1. I think the institution involved gets it’s claws on the copyright for most things and it’s either too much BS to get released or they actually say no, we want money for it and try to pawn it off to potential interested parties.

      Believe it or not the Berkeley licence was seen as a huge step forward compared to most .edu attitudes.

    1. Would electroplating create a decent antenna? I saw the article on this several days ago and thought about metallic paint. Looked it up and some guys doing radio/dish antennas found that the paint was a good deal less effective than solid metal. Maybe electroplating will fair better.

    1. I use aluminum duct tape on my generic reflectors. I like to smooth out the best I can. I’ve used craft copper and conductive copper tape also for antennas and shielding though not reflectors. Passive director elements though.

  4. But doesn’t this essentially break the MiMo and beam forming design of these multiple antenna routers? It looks like you’re taking something that could have used MiMo and beam forming to increase bandwidth and steer the beam in the direction of the client into a sectorized antenna AP.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s