Wireless USB + UWB Demotion Goes Ahead For Linux 5.4

[phoronix]

Phoronix: Wireless USB + UWB Demotion Goes Ahead For Linux 5.4

Back in August I noted that Wireless USB and Ultra Wideband would be deprecated within the Linux kernel and that is indeed happening for Linux 5.4...

http://www.phoronix.com/scan.php?pag...cates-WUSB-UWB

[OlafLostViking]

This sounds quite unfortunate, taking into account that some industries just started to recognize UWB as feasible technology (see UWB Alliance, FiRa Consortium). Beginning with next year VW together with NXP wants to use UWB for precise measurements and access control for their new cars.

I hope that some of these many companies beginning to use UWB will take over the development of the code so that this technology stays/becomes accessible for the hacker community when it gets introduced into our lives via cars or smart homes etc.

[wizard69]

This sounds quite unfortunate, taking into account that some industries just started to recognize UWB as feasible technology (see UWB Alliance, FiRa Consortium). Beginning with next year VW together with NXP wants to use UWB for precise measurements and access control for their new cars.

I hope that some of these many companies beginning to use UWB will take over the development of the code so that this technology stays/becomes accessible for the hacker community when it gets introduced into our lives via cars or smart homes etc.

Maybe they will take on a complete rewrite? One is left with the feeling that UWB is just starting to get some acceptance when the Linux world is about to demote.

[[email protected]]

They are demoting it because nobody care. IF it start to be a thing (it never was), there is no obstacle bringing it back to form, as long as somebody take responsibility for it.

[alcalde]

They are demoting it because nobody care. IF it start to be a thing (it never was), there is no obstacle bringing it back to form, as long as somebody take responsibility for it.

It was the most amazing thing in the whole world.

Ultra-wideband is also used in "see-through-the-wall" precision radar-imaging technology, precision locating and tracking (using distance measurements between radios), and precision time-of-arrival-based localization approaches. It is efficient, with a spatial capacity of approximately 10¹³ bit/s/m². UWB radar has been proposed as the active sensor component in an Automatic Target Recognition application, designed to detect humans or objects that have fallen onto subway tracks.

.... In terms of military use, a UWB Doppler radar could demonstrate ground, foliage, and wall penetrating capabilities.

Not my fault no one took the trouble to show how a Raspberry Pi, Linux and probably 10 lines of Python code could let you see through walls.

In some sort of weird Steisand-effect thing, the announcement of the deprecation is the first most people have heard of UWB and thus they're only now realizing how awesome it is.

[tg--]

It was the most amazing thing in the whole world.

At best it was an interesting idea.

Not my fault no one took the trouble to show how a Raspberry Pi, Linux and probably 10 lines of Python code could let you see through walls.

UWB, as planned for end-user devices would in no sense be usable as radar technology.
It was basically only going to be available in the 8-10 GHz range, at transmit power levels of around -40 dBm/MHz (and well below -60 dBm/MHz for other ranges), which means it basically doesn't go through walls and certainly can't bounce back from objects behind those walls.

There were UWB-based radar systems centered at around 24 GHz (with 5 GHz bandwidths), mainly thought as very short range radar for cars.
But that's being phased out, the 24 GHz band is no longer available for UWB in many parts of the world, and the system wasn't a success.
In this band lies a strong H2O absorption-band, which means those radars work poorly in rainy and foggy weather.

UWB also has undesirable side-effects.
It raises the noise level in a very large frequency range (you might call it an ultra-wide noise source), interfering with all other devices operating in that band, and interfering heavily with radio-astronomical observation stations.

The system also uses a lot of power making mobile UWB-devices bulky and expensive, which is the main reason UWB has died (and yes, it has).

Apart from these technology-specific issues, the Linux UWB support is also problematic.
It doesn't actually implement much of UWB, it only implements the basic device support infrastructure for WUSB (which is long dead).
The Linux UWB stack would in no way be usable for implementing a radar solution, it can only set up devices for WUSB.
It in no way implements any radio-support, only basic channel selection - again for WUSB.

Long story short:
WUSB is dead, nobody wants it, and nobody works on it. Almost no devices exist, no new ones will ever come.
UWB is also dead, there are other radio technologies that replaced it long ago, and even for the few things it was useful for, it is being replaced by other technology.
Most of the worlds regulators dropped it, no new UWB hardware is being developed.
WUSB and UWB are dead, and the linux support is a relic of a failed technology from a past decade.

[mikus]

The new iphones use uwb suddenly for their airdrop to speed up random transfers from strangers. Maybe not as dead as it could be since it might be fun to explore from a security perspective.

[szymon_g]

sad news to both its users...

[mlau]

At best it was an interesting idea.
UWB, as planned for end-user devices would in no sense be usable as radar technology.

I was told at university 20 years ago by an Infineon engineer that the basic idea of UWB is to spread the signal power over a very wide band (few hundred MHz to GHz),
with a special spreading code. So instead of having a power peak around the center frequency it was basically "hiding" the signal in the background noise. The same spreading code
applied in the receiver would then filter out the signal from the noise. I always thought it was a brilliant technology for clandestine transmissions: When you casually look at
a signal plot you wouldn't notice anything because it's indistinguishable from the noise floor.

[starshipeleven]

I was told at university 20 years ago by an Infineon engineer that the basic idea of UWB is to spread the signal power over a very wide band (few hundred MHz to GHz),
with a special spreading code. So instead of having a power peak around the center frequency it was basically "hiding" the signal in the background noise. The same spreading code
applied in the receiver would then filter out the signal from the noise. I always thought it was a brilliant technology for clandestine transmissions: When you casually look at
a signal plot you wouldn't notice anything because it's indistinguishable from the noise floor.

That kind of transmission will need some serious ECC algorithms, as if it's at the same power level of the background noise (thus indistinguishable from it) then it's going to be often overridden by it and this means you lose the data in the parts where the signal was overridden.