By Francis VaughanPosted Friday 5th September 2008 13:52 GMT
A very neat solution.
And no - arrhythmia isn't a problem. What they have done is to solve the key distribution problem. Here the key is encoded the time between a series of heartbeats. They showed that, in principle, the pacemaker can observe the heartbeat intervals, and the external programmer can observe the same intervals. They both see the same data, and hey presto - shared key. What you are hoping for is a reasonable amount of entropy in that shared observation. Enough to find 64 bits (at least) of shared key.
There are lots of different cardiac conditions, and as many pacemaker types. Even those conditions that require every heartbeat to be triggered by the pacemaker could in principle include enough variation in period to encode the key - in this case the pacemaker itself would effectively be generating and transmitting the key in its normal operation. Pacemakers that watch and only step in when needed would rely on the heart's natural jitter in heartbeat period to generate the key. The worst thing would be a very stable heartbeat - since there would not be so much entropy in the shared key.
By Anonymous CowardPosted Friday 5th September 2008 14:21 GMT
This is all good and well, but this helps proove the point that biometrics is a bad form of authentication because it's non-revokable. At the base level its security by obscurity. Hide the pattern and hope to god no-one else finds it and can replicate it.
So when fingerprints and iris scanners are common place, lets move onto anatomical scans (Heartrate, size of forearm, how many hairs on your deriarse?) but its always going to get easier to scan for that info. Tit for Tat.
On the otherhand, a fast processor can break a private key. So keep increasing it, bigger harder numbers, New algorithms. Ah, Quantum computing :- there is quantum cryptography too. PKI ensures that the probability of a identity theft is very low, assuming that the private key is kept secret. Just as secret as your fingerprint or the formula for predicting your heartrate.
The point is that if anyone discovers your private key, you change it. If anyone discovers the blueprints to your body.. what? Kill yourself?
Meanwhile its one hell of a money feast at the public pork barrel.
By Anonymous CowardPosted Friday 5th September 2008 14:30 GMT
That'd be fine. The system only allows you to fiddle with the settings if the remote has a matching signal. A doctor would have the remote hooked up to the patient, but Boris T Hacker won't.
By ThomasPosted Friday 5th September 2008 14:44 GMT
My reading of the article is that you connect a device to the patient's finger, it triggers the implant to prepare a key, both devices watch the next 16 beats to build a key, then use it for secure communications. So how are arrhythmia or heart transplants a problem?
By yeah, right.Posted Friday 5th September 2008 16:12 GMT
So, let me get this straight. In order to communicate with a wireless device, they need to create a physical connection to the person carrying said device?
Wouldn't it just be easier to have some sort of "through the skin" induction-like direct connection and get rid of the wireless completely? Benefit there is increased battery life as it then doesn't have to drive that signal.
By Kevin McMurtriePosted Friday 5th September 2008 17:59 GMT
Does no heartbeat map to a default key as a safety backup? I'm working on plans for my remote controlled zombie army.
heartbeat pressure waves travelling through elastic arteries... #
By John BensonPosted Friday 5th September 2008 19:56 GMT
...are analogous to electrical pulses travelling through transmission lines with capacitance: waveforms will deform.
Blood vessels will constrict to different degrees at different times (analogous to changing the capacitance in an electric line), so the agreement between the two heartbeat waveforms can be expected to fluctuate over time. Whether the signal degradation would be enough to affect the timing scheme is open to question. I just hope they're testing over a wide range of constriction.
By Solomon GrundyPosted Friday 5th September 2008 23:15 GMT
I'm not the best at maths, but it seems to me that there just aren't enough variables in a heartbeat to make this too challenging to overcome. If someone could clarify for me it'd be great.
By Allan DyerPosted Saturday 6th September 2008 04:22 GMT
So, if you can monitor someone's heartbeat, perhaps when they touch something, e.g. a door handle, you can take control of their implants.
On the other hand, wiring the door handle to a high voltage supply is probably a more effective, though less subtle means of attack. And you don't have to wait for your target to hold the handle for 16 heartbeats.
By David WilkinsonPosted Saturday 6th September 2008 16:44 GMT
The devices would only authenticate if both are taking the same reading from the patient at the same time.
Prevents someone from committing random murders by hacking together some gear and wirelessly reprogramming people's pacemakers.
Then again I am hoping current pacemakers do some sanity checking in the range of values they accept, and that the wrong values wouldn't lead to sudden death.
By Daniel GarciaPosted Saturday 6th September 2008 18:27 GMT
well, the variables that we take in consideration are the variation of the size between the heartbit duration.
made up example:
using the first interval like unit and then you compare with the others intervals, for example. if you measure with accuracy of 1000th parts, then every interval after first heart bit gives you to at least 3 decimal, discard the first decimal( as will be commonly 0,1 or 9) and the 3rd( as will contain the measurement error),taking the second and "more" ramdon cipher, you got at least 14 digits decimal number like:
58692834495220 using an approx 10^3 ~= 2^8 this is on the 2^(8*5) range, or 40 bits.
the key size depens on the number of heart bit measure and the accuracy of measurement and the number of cipher used per measurement.
By Anonymous CowardPosted Sunday 7th September 2008 02:45 GMT
Yea - review folks, it doesn't matter if you have a baboon's heart - the heart is simply the generator of the common key both "listeners" need to chat. Heck, you could probably use a microphone on the gurgles of stomach acid for 15secs to generate a sufficient key.
Just comes down to true strong authentication: Have something you know + something you have/are/do... in this case, you have to know you need 2nd listening device and the something you have is the heart's natural rhythm (so long as the pace maker isn't active the entire time they are listening...)
But really folks - i know it's academic, bet we're talking about people hacking a sick person's heart - so Rube Goldberg... just freaking feed the person McD's double quarter pounders or maybe that new "Baconator", or even better - fire up a degausser... - they won't be 'round much longer after that. >:-)
By Anonymous CowardPosted Monday 8th September 2008 08:44 GMT
umm didn't they already do this on mission impossible 3?
p.s i have it on good authority that the record of accidental defibrillations at the top heart hospital currently stands at 32.. (making my paltry 8 seem a little febrile)
Comments on: Boffins use heartbeat to thwart wireless implant hack
This is wonderful news for patients... #
By KnownCoward Posted Friday 5th September 2008 12:54 GMT
Less wonderful news #
By Anonymous Coward Posted Friday 5th September 2008 13:10 GMT
Seems to me... #
By Anonymous Coward Posted Friday 5th September 2008 13:44 GMT
Very cute #
By Francis Vaughan Posted Friday 5th September 2008 13:52 GMT
@KnownCoward #
By Knasher Posted Friday 5th September 2008 13:56 GMT
@arrhythmia. #
By Anonymous Coward Posted Friday 5th September 2008 14:12 GMT
Biometric Security #
By Anonymous Coward Posted Friday 5th September 2008 14:21 GMT
What if their heart has stopped.... #
By Anonymous Coward Posted Friday 5th September 2008 14:25 GMT
@Known Coward. #
By Anonymous Coward Posted Friday 5th September 2008 14:30 GMT
Re: Less wonderful news #
By Anonymous Coward Posted Friday 5th September 2008 14:38 GMT
What could possibly go wrong??? #
By Timo Posted Friday 5th September 2008 14:41 GMT
What? #
By Thomas Posted Friday 5th September 2008 14:44 GMT
connection? #
By yeah, right. Posted Friday 5th September 2008 16:12 GMT
Ack... #
By Kevin McMurtrie Posted Friday 5th September 2008 17:59 GMT
heartbeat pressure waves travelling through elastic arteries... #
By John Benson Posted Friday 5th September 2008 19:56 GMT
Not a Maths Whiz #
By Solomon Grundy Posted Friday 5th September 2008 23:15 GMT
Beware of door handles... #
By Allan Dyer Posted Saturday 6th September 2008 04:22 GMT
@What if their heart has stopped.... #
By Doug Glass Posted Saturday 6th September 2008 09:10 GMT
am I missing something ? #
By tony trolle Posted Saturday 6th September 2008 16:04 GMT
It does the matching in real time. #
By David Wilkinson Posted Saturday 6th September 2008 16:44 GMT
RE:Not a Maths Whiz #
By Daniel Garcia Posted Saturday 6th September 2008 18:27 GMT
Uhh #
By Anonymous Coward Posted Sunday 7th September 2008 02:45 GMT
global stupidity season #
By Anonymous Coward Posted Monday 8th September 2008 08:37 GMT
hold on a doggone minute #
By Anonymous Coward Posted Monday 8th September 2008 08:44 GMT
Dead on #
By rick buck Posted Tuesday 9th September 2008 06:08 GMT
thinks again #
By tony trolle Posted Wednesday 10th September 2008 06:15 GMT