MCWKen
Member
As far as I am aware of, there is no radio or software that can decode AES-256 encryption. There are some who claimed they have done it, don't believe them (including the Russians and North Koreans).
Polk Co question on SARA
- Thread starter mws72
- Start date
I've found that many end users conflate P25 Digital with Encrypted. There were many departments around me that thought that switching to ISICS was going to eliminate the possibility of scanner listeners monitoring traffic.
Yes, Phase 2 scanners are expensive, but they can't decode encryption. Could someone possibly decrypt a P25 transmission? Sure, but it isn't going to be done on the fly without the firmware being designed to do so and having access to the appropriate key. Certainly not in real time.
Yes, Phase 2 scanners are expensive, but they can't decode encryption. Could someone possibly decrypt a P25 transmission? Sure, but it isn't going to be done on the fly without the firmware being designed to do so and having access to the appropriate key. Certainly not in real time.
The LEO doesn't understand the technology - not surprising. There is no radio on the market that can decrypt (not decipher) encrypted communications without having the proper keys installed.
bobcraig
Member
That's what I thought. Thanks!
Lets just say that you received a message encrypted in AES-256. The entire key needs to be brute forced with a string of 256 bits that can either be 1 or 0. That makes the number of possibilities 2 to the power of 256 or 115 quattrovigintillion combinations. (115,792,089,237,316,195,423,570,985,008,687,907,853,269,984,665,640,564,039,457,584,007,913,129,639,936).
The most reasonable hardware to accomplish this would be a high end GPU (RTX 4090) which can theoretically compute 82.58 teraflops. Just for simplicity sake, lets round that up to 100 TFLOPS.
Let's say you have 1 billion GPU's strung together in one machine capable of attempting to brute force the encryption key. That will get 100 million petaflops. Now, there's 31,556,952 seconds in a year. So a billion of these GPU's that don't technically even exist yet could theoretically brute force 3,168 keys per second which means it would take 3.628905e+16 (36,289,050,000,000,000) years to decode.
That's over 36 quadrillion years to brute force the key. Theoretically it should be broken before getting halfway through, so lets say 18 quadrillion years. The universe's age by best estimates is only 14 billion years old so it would take 1,285,714 times longer than the universe has existed to try half of the available keys using theoretical hardware that isn't available to the consumer yet.
And that's not even talking about the amount of power it would take to decrypt that. A RTX4090 has a TDP of 450 watts. Even if a theoretical device could muster the same compute power at 100 watts, we have 1 billion of them working on this, that's 100 billion watts of energy or 100 gigawatts. The average monthly power consumption of the entire United States is 877MW meaning that it would require more than 9 years of average energy consumption of the entire United States to theoretically brute force half of the possibilities for the key. The most powerful nuclear reactor (Kashiwazaki-Kariwa) generates 1GW of power meaning that there would need to be more than 100 of the most powerful nuclear reactors to power such a farm.
But 1 billion GPU's is pretty unrealistic... So, how about a supercomputer?
The Tianhe-2 Supercomputer can process 33.86 petaflops. It would still take more than 54 septendecillion years. That's just for one machine. Reducing the time by just one power would require 10 more additional basketball court sized supercomputers.
I have given some rather generous allowances in these calculations, and I'm not a mathematician, just a civil servant, so take that with a grain of salt.
So, it's not impossible to brute force an AES-256 key... but highly unlikely... Not even for China or Russia.
The most reasonable hardware to accomplish this would be a high end GPU (RTX 4090) which can theoretically compute 82.58 teraflops. Just for simplicity sake, lets round that up to 100 TFLOPS.
Let's say you have 1 billion GPU's strung together in one machine capable of attempting to brute force the encryption key. That will get 100 million petaflops. Now, there's 31,556,952 seconds in a year. So a billion of these GPU's that don't technically even exist yet could theoretically brute force 3,168 keys per second which means it would take 3.628905e+16 (36,289,050,000,000,000) years to decode.
That's over 36 quadrillion years to brute force the key. Theoretically it should be broken before getting halfway through, so lets say 18 quadrillion years. The universe's age by best estimates is only 14 billion years old so it would take 1,285,714 times longer than the universe has existed to try half of the available keys using theoretical hardware that isn't available to the consumer yet.
And that's not even talking about the amount of power it would take to decrypt that. A RTX4090 has a TDP of 450 watts. Even if a theoretical device could muster the same compute power at 100 watts, we have 1 billion of them working on this, that's 100 billion watts of energy or 100 gigawatts. The average monthly power consumption of the entire United States is 877MW meaning that it would require more than 9 years of average energy consumption of the entire United States to theoretically brute force half of the possibilities for the key. The most powerful nuclear reactor (Kashiwazaki-Kariwa) generates 1GW of power meaning that there would need to be more than 100 of the most powerful nuclear reactors to power such a farm.
But 1 billion GPU's is pretty unrealistic... So, how about a supercomputer?
The Tianhe-2 Supercomputer can process 33.86 petaflops. It would still take more than 54 septendecillion years. That's just for one machine. Reducing the time by just one power would require 10 more additional basketball court sized supercomputers.
I have given some rather generous allowances in these calculations, and I'm not a mathematician, just a civil servant, so take that with a grain of salt.
So, it's not impossible to brute force an AES-256 key... but highly unlikely... Not even for China or Russia.
Any P25 radio with an AES encryption feature can 'decipher' encryption. However not only is this illegal without authorization but you would need the encryption keys loaded in to the radio to do so, which can only be obtained from the system administrator. What he was likely referring to is the radios that a majority of sara users in Polk county uses, the Harris XL series. These can be found used on ebay for usually around $3-4,000. Even if you were to purchase one, encryption keys are not given out to anyone not authorized to be on the radio system, and AES encryption is essentially impossible to crack.
MCWKen
Member
Well.......that leaves North Korea out. LOL
newsphotog
Member
Westcom will be testing encryption on the info channel and some of the tac channels starting next year. And it's not just you... some of the officers have a hard time hearing some of the dispatchers too. Also scanners are not the best at decoding digital for the best audio. It sounds way better on a purpose-built P25 radio than it does on a scanner.
bobcraig
Member
Interesting..... Thanks!
This is some good information. I was wondering why I was not hearing the Ankeny PD anymore.
Similar threads
- Locked
