Can a microprocessor work as a reciever?

A place to discuss the science of computers and programs, from algorithms to computability.

Formal proofs preferred.

Moderators: phlip, Moderators General, Prelates

User avatar
david.windsor
Posts: 121
Joined: Mon Sep 09, 2013 3:08 pm UTC

Can a microprocessor work as a reciever?

Postby david.windsor » Thu Aug 27, 2015 7:47 pm UTC

Can a microprocessor work as a reciever? We can read the EMI a processor transmits and read what state it registers are in; can strong enough, complex enough EMI do the reverse and set the registers to a desiered state?
"All those ... moments, will be lost ... in time, like tears ... in rain."

wumpus
Posts: 494
Joined: Thu Feb 21, 2008 12:16 am UTC

Re: Can a microprocessor work as a reciever?

Postby wumpus » Fri Aug 28, 2015 2:51 pm UTC

I'm pretty sure the answer is yes, but only under certain conditions (have fun setting the program counter).

First: I'm pretty sure your assumptions are incorrect. While you can determine a great deal about the state of a CPU/computer by analyzing the EMI emissions, I doubt you can accurately determine specific bits without bouncing emissions from a [scannning] electron microscope over them.

Second: The above means (or simply blasting a single source of EMI at a CPU) will only set bits in one direction. You need to set all (or all that matter, still a huge number) of the bits simultaneously.

Third: the minimum trace size (the xx in xxnm) is less than the wavelength of light, much less any more controllable form of EMI. While you are trying to set a 6 transistor (or more) register, you still will be using what is commonly called light. My guess is that hitting it hard enough with light will set/clear the register, but you won't really have a means to flip it the other way. You might be able to control *much* older chips with EMI you can directly modulate.

Fourth: your "laser blaster" would look almost identical to a modern fab processs (ok, significantly out of date since you only need to hit transistors instead of transistor parts, but in general a fab). You are blasting light through an extremely fine mask just like a fab. Expect the price to build the equipment to be close to a billion dollars.

I think that pretty much covers it. For a billion dollars, maybe you can flip the bits you need in a certain direction. For much older chips, you might be able to rig up a maser holograph that will work on chips that still have the lids, but don't expect it to be much cheaper.

PS. This is literally how EPROMS used to be erased. You hit them with EMI (UV light) and got them to the desired state (either all zeros or all ones, I forgot which. This means if you needed to flip a bit in one direction you still could, but going the other way meant pulling it an putting it under the UV light).

User avatar
david.windsor
Posts: 121
Joined: Mon Sep 09, 2013 3:08 pm UTC

Re: Can a microprocessor work as a reciever?

Postby david.windsor » Fri Aug 28, 2015 5:20 pm UTC

oh well there goes my great idea for hi-jacking drones. :lol:
"All those ... moments, will be lost ... in time, like tears ... in rain."

Tyndmyr
Posts: 10130
Joined: Wed Jul 25, 2012 8:38 pm UTC

Re: Can a microprocessor work as a reciever?

Postby Tyndmyr » Wed Sep 09, 2015 7:49 pm UTC

wumpus wrote:PS. This is literally how EPROMS used to be erased. You hit them with EMI (UV light) and got them to the desired state (either all zeros or all ones, I forgot which. This means if you needed to flip a bit in one direction you still could, but going the other way meant pulling it an putting it under the UV light).


Light sensitivity still shows up, some places. If memory serves, the raspberry pi 2 has a reboot issue with sufficient light.

As for jacking drones, well...capturing drones shouldn't actually be that hard if one wanted to do so. Waaaay less technical aptitude is needed. Net guns for everyone.

User avatar
david.windsor
Posts: 121
Joined: Mon Sep 09, 2013 3:08 pm UTC

Re: Can a microprocessor work as a reciever?

Postby david.windsor » Wed Sep 09, 2015 11:39 pm UTC

I like net guns, but how cool would it be to fly a drone up next the target, reprogram it and add it to your growing hive mind of drones.
"All those ... moments, will be lost ... in time, like tears ... in rain."

Jeff_UK
Posts: 66
Joined: Sat Nov 03, 2007 10:38 pm UTC

Re: Can a microprocessor work as a reciever?

Postby Jeff_UK » Tue Dec 15, 2015 5:02 pm UTC

david.windsor wrote:I like net guns, but how cool would it be to fly a drone up next the target, reprogram it and add it to your growing hive mind of drones.


You could try GPS spoofing, fly next to them and transmit GPS data at a much higher power than the satellites can muster; You could very easily control them using an algorithm like:

Determine your desired instruction (move at velocity X)
Identify current instruction (watch which way it's trying to move, call this velocity Y)
Provide fake GPS information such that velocity Y, relative the fake signal, is velocity X relative to the real signal)

Rinse and Repeat...

You could control multiple drones by determining the frequency at which they check for GPS signal, and rotating through instructions in line with each of your targets; they might be a bit jerky...

Edit: someone beat me to it.. http://hackaday.com/2012/07/01/spoofing ... r-own-uav/


Or, you could just use a net.
"Please only print this post if you really need to"
...hmm....I wonder how much extra energy is required to generate that request...We need a cost/benefit analysis, STAT!


Return to “Computer Science”

Who is online

Users browsing this forum: No registered users and 6 guests