No, this is to concentrate the radio waves into a desired shape. Stronger in this direction, weaker than that.
Think of it very vaguely like a parabolic mirror on a flashlight directing the light vs a naked light bulb putting light out in all directions. (this is a bad metaphor for what's going on but it's the basic idea of the goal)
To change the direction you have to physically move the antenna OR have an active phased array antenna with an electronic component which has a variable phase change to be able to move the beam around while leaving the antenna fixed.
It does do directional steering. No active array components or physical movement are needed.
> If the output ports are connected to individual antennas in an antenna array, this allows shaping the beam in different directions by switching which input port the signal is sent to.
From TFA.
Presumably the geometrical shape of the lens is dictated by solving for useful phase shifts for different input points. Otherwise you could just use a bunch of delay lines.
I wonder if anybody's ever designed a 3D version of this. You might get a wider range of inputs, or more precise steering, by shaping the delays on a non-Euclidean (curved) surface (like a sphere or a saddle).
>It does do directional steering. No active array components or physical movement are needed.
By... plugging and unplugging antenna elements?
It makes a fixed directional antenna element array which is configurable to a small degree by choosing which antenna elements to connect and their spatial arrangement.
The radiation pattern can only change by physically plugging or moving antenna elements.
No. The antenna elements are always plugged in the same way. By sampling/introducing the signal at a different physical position on the lens (or multiple positions simultaneously), you create different physical distances and phase shifts for each element, and therefore a different beam direction.
...Apropos of nothing, this reminds me that a long ago, I played on a Minecraft server where one of the boys made a piston display with controlled by a pressure plate array. I was shocked by how he managed to transmit the signal with a simple wire. Maybe it was a similar idea, using the propagation distance...
>By sampling/introducing the signal at a different physical position on the lens
Which is accomplished with active components (or multiple frontends). The word "active" is carrying a lot of weight in what isn't actually that important of a distinction
You can aim an aerial by physically rotating it. You've probably done this, gone up on the roof to adjust the aim of your TV aerial or satellite dish. It makes sense, right? A Yagi aerial - a reflector, a driven element, and a bunch of directors - focuses the beam in a kind of aubergine-shaped blob in the direction of the pointy end.
But you can also aim aerials by having two of them, and varying the phase that you send a signal into them. This sounds a bit mental but consider how direction finding equipment like LoJack works - you have a transmitter in an unknown location and you have a cluster of aerials connected to one receiver. By comparing the phase of the incoming signal between two aerials you can work out which one it's nearer to! This trick works well enough if you make two dipoles spaced a half wavelength apart that you can easily homebrew something where by switching in a 180° phase shift at an audio rate, the difference in phase can be heard as a tone.
In this case you've got a bunch of aerials attached to the ports along the bottom and the phase of the signal reaching them depends on how long it's taken to cross the microstrip. If you fire it in at the top in the centre it'll be equal (notice the middle "legs" have kinks in them to keep the path length the same?), if you fire it in at the side then one of the ports at the side will get the signal sooner and its phase will appear advanced compared to the other one - and the beam will bend that way.
Imagine a power splitter + phase shifter that produces a correct phase shift for each element in a phased array to produce a directional beam from one radio transceiver.
Now this clever arrangement, instead of having only one radio transceiver port, has multiple. And each of those ports corresponds to a different set of phase shifts, producing a directional beam at a different azimuth angle.
And because this is an entirely passive device, it's linear, and all ports can be active at the same time (principle of superposition essentially). So you can use a single phased antenna array to serve multiple directional beams at the same time.
Basically, it lets you steer a beam very precisely in an arbitrary direction, as if you were physically rotating the transmitting antenna.
Usually you'd need to rotate the antenna or else use a large number of controllable phase-shifting elements to send an output beam in the desired direction. But a Rotman lens lets you use a small number of phase shifting elements to merge a small number of source beams (still > 1, though) into a single precisely-steered output beam.
If you monitored all the inputs simultaneously instead of switching, you could make a low-tech radio-wavelength camera. Presumably with less SNR per "pixel" than you'd get from monitoring just one input though.
It's a way to create constructive interference in an RF signal in a certain direction (the signal gets stronger in one direction and weaker in others without changing the hardware). It's commonly used in LTE and Wi-Fi as a way to increase SNR directionally for clients.
Re:Rotman Lens
by bt1a on Monday July 06, 2026 @05:11PM
(#48810571)
I was expecting some sort of fungal network with measurable compute on initial glance. Then the article image couldn't help but look quite like the outline of an Earthbound entity.
Oh I dunno. I made a mm-wave radar with a Rotman lens using a generative loop between Python, Rhino, and EM simulation. Pretty sure AI could cook that up.
Used to work in photonics. The star coupler portion is a Rotman lens, the whole thing is different (an AWG). It's used to spectrally separate an input (or in this case, inputs) across different outputs.
Tito Beveridge creator of Tito Vodka, Ed Currie creator of the Carolina Reaper and Pepper X, Keith Weed the president of Britain’s Royal Horticultural Society, Usain Bolt one of the quickest people in the world, Thomas Crapper - commonly but incorrectly attributed with the invention of the modern toilet - founder of Crapper & Co plumbing company, meteorologist Sarah Blizzard, LEO Rob Banks, Donna/Mike/Bobbie Singer all singers with no relation.
I myself have a CTO of a plumbing company in my family from Flushing MI, but that's ever so slightly different.
Re:Rotman Lens
by Tade0 on Tuesday July 07, 2026 @06:57AM
(#48816065)
I think nominative determinism has it backwards: people are actually nudged into careers which fit their names as whoever is on the receiving end of their services automatically has a more positive sentiment if they notice the name fits.
I experienced this myself as my first name was rare in my generation and normally found in history books or literary classics so I was perceived through that lens.
Re:Rotman Lens
by bbminner on Monday July 06, 2026 @10:29PM
(#48813028)
How difficult it is to get a perfect beam in practice? Can one manufacture a flat rotman lense like that to introduce sinc phase shifts - this should yield a well behaved very directed beam? What are the practical challenges in doing so?
Suppose this is connected to a multiple-output RF transmitter (amplifier), what are the advantages/applications of shifting the phases of the signals passively with a Rotman Lens versus shifting the phases of the signals prior to feeding them to the transmitter (i.e. in the generator)? is the main purpose to compensate for different antennae geometry without having to change the signal generator? I guess things are very different when you can't commute everything into your SDR algorithms.
Re:Rotman Lens
by Skwid on Tuesday July 07, 2026 @07:02AM
(#48816105)
You can do something similar in 3D with Luneberg lenses, as used on the AN/SPG-59[1]. A truly wild piece of engineering, as is the following SCANFAR system if you're into that sort of thing.
I can only assume someone in the top brass had a skim of Arthur C Clarke's Superiority and forwarded it straight on to engineering with a big green approved stamp.
Re:Rotman Lens
by amarant on Tuesday July 07, 2026 @07:58AM
(#48816508)
R.F Turner is a pretty nice example of nominative determinism. Parents up and named him "radio frequency dial", so of course he works in antenna technology
Think of it very vaguely like a parabolic mirror on a flashlight directing the light vs a naked light bulb putting light out in all directions. (this is a bad metaphor for what's going on but it's the basic idea of the goal)
To change the direction you have to physically move the antenna OR have an active phased array antenna with an electronic component which has a variable phase change to be able to move the beam around while leaving the antenna fixed.
> If the output ports are connected to individual antennas in an antenna array, this allows shaping the beam in different directions by switching which input port the signal is sent to.
From TFA.
Presumably the geometrical shape of the lens is dictated by solving for useful phase shifts for different input points. Otherwise you could just use a bunch of delay lines.
I wonder if anybody's ever designed a 3D version of this. You might get a wider range of inputs, or more precise steering, by shaping the delays on a non-Euclidean (curved) surface (like a sphere or a saddle).
By... plugging and unplugging antenna elements?
It makes a fixed directional antenna element array which is configurable to a small degree by choosing which antenna elements to connect and their spatial arrangement.
The radiation pattern can only change by physically plugging or moving antenna elements.
No. The antenna elements are always plugged in the same way. By sampling/introducing the signal at a different physical position on the lens (or multiple positions simultaneously), you create different physical distances and phase shifts for each element, and therefore a different beam direction.
...Apropos of nothing, this reminds me that a long ago, I played on a Minecraft server where one of the boys made a piston display with controlled by a pressure plate array. I was shocked by how he managed to transmit the signal with a simple wire. Maybe it was a similar idea, using the propagation distance...
Which is accomplished with active components (or multiple frontends). The word "active" is carrying a lot of weight in what isn't actually that important of a distinction
You can aim an aerial by physically rotating it. You've probably done this, gone up on the roof to adjust the aim of your TV aerial or satellite dish. It makes sense, right? A Yagi aerial - a reflector, a driven element, and a bunch of directors - focuses the beam in a kind of aubergine-shaped blob in the direction of the pointy end.
But you can also aim aerials by having two of them, and varying the phase that you send a signal into them. This sounds a bit mental but consider how direction finding equipment like LoJack works - you have a transmitter in an unknown location and you have a cluster of aerials connected to one receiver. By comparing the phase of the incoming signal between two aerials you can work out which one it's nearer to! This trick works well enough if you make two dipoles spaced a half wavelength apart that you can easily homebrew something where by switching in a 180° phase shift at an audio rate, the difference in phase can be heard as a tone.
In this case you've got a bunch of aerials attached to the ports along the bottom and the phase of the signal reaching them depends on how long it's taken to cross the microstrip. If you fire it in at the top in the centre it'll be equal (notice the middle "legs" have kinks in them to keep the path length the same?), if you fire it in at the side then one of the ports at the side will get the signal sooner and its phase will appear advanced compared to the other one - and the beam will bend that way.
Now this clever arrangement, instead of having only one radio transceiver port, has multiple. And each of those ports corresponds to a different set of phase shifts, producing a directional beam at a different azimuth angle.
And because this is an entirely passive device, it's linear, and all ports can be active at the same time (principle of superposition essentially). So you can use a single phased antenna array to serve multiple directional beams at the same time.
Usually you'd need to rotate the antenna or else use a large number of controllable phase-shifting elements to send an output beam in the desired direction. But a Rotman lens lets you use a small number of phase shifting elements to merge a small number of source beams (still > 1, though) into a single precisely-steered output beam.
https://en.wikipedia.org/wiki/Beamforming
https://academy.lucedaphotonics.com/modules/awg/guide/compon...
Astonishing nominative determinism there
I myself have a CTO of a plumbing company in my family from Flushing MI, but that's ever so slightly different.
Amy Lloyd, British neuroscientist who studied amyloid plaques associated with Alzheimer's disease[0]
[0] https://www.dundee.ac.uk/people/amy-lloyd
I experienced this myself as my first name was rare in my generation and normally found in history books or literary classics so I was perceived through that lens.
Some studies have shown statistically significant effects from having aptonyms. More than the expected number of dentists named Payne, etc.
Any open source packages that could do it?
1. https://secwww.jhuapl.edu/techdigest/content/techdigest/pdf/...