3 ms·
During WWII, when the British succeeded in making their territory spy-free, (and convinced German spies to co-operate) they got every single new spy sent acros
by DoctorOetker 1y ago
During WWII, when the British succeeded in making their territory spy-free, (and convinced German spies to co-operate) they got every single new spy sent across the channel and forced them into compliance as well. They would send back false "warmer / colder" (V1 / V2) rocket impact site locations to Germany, who believed in their compromised spy network. This allowed the British to redirect the damage to less critical target areas.
If the West is indirectly delivering components, or even microcomputers to Russia, couldn't they insert small extra sub-circuits in the silicon which detects by connection topology its in a Russian drone? Depending on this detection it could plausibly be made to operate properly on (Bela)Russian occupied soil (so it isn't noticed during testing etc.), but then malfunction on non (Bela)Russian occupied soil?
The first versions could just dump the RAM / flash contents and other state somewhere (say secret extra Flash memory on the die, which might be small even, and then just randomly select a source address range, so that multiple downed or recovered drones eventually provide a full image).
Once one has the assembly / Forth / whatever software stack available for analysis, the next iteration of compromised microcomputers could be made to more intelligently make active use to say safely land, or even return to source etc midair.
Every drone that "spontaneously" safely lands on any designated areas when in actual use (as opposed to during testing etc.), saves a lot of money for rockets etc. to disable it.