GSM encryption cracked, showing its age
- — 30 December, 2009 06:12
The unveiling of a GSM (Global System for Mobile Communications) encryption codebook compiled by a German security researcher and his team of collaborators lowers the bar significantly for the amount of money and technical expertise required to listen in on a GSM-based mobile phone call. More importantly, it illustrates just how old the current GSM encryption is and demonstrates why it's time for an upgrade.
Law enforcement officials and well-financed cyber criminals have been able to crack GSM encryption for some time, but the investment was so high that it didn't pose much of a threat. This new method lowers the price of entry to the point that it is more of an issue, but still not a high risk.
Karsten Nohl announced that he and his team have compiled 2 terabytes worth of GSM encryption data. PC World's Robert McMillan explains that the results are like "cracking tables that can be used as a kind of reverse phone-book to determine the encryption key used to secure a GSM (Global System for Mobile communications) telephone conversation or text message."
GSM is the most widely-used mobile phone technology in the world — accounting for over 80 percent of the world's 4.3 billion mobile phones. The encryption algorithm that protects GSM-based calls from being intercepted and eavesdropped is more than twenty years old, though.
Time is the enemy of encryption. When a new encryption algorithm is developed and claimed to be impenetrable, or that cracking it is so impractical as to not be plausible, those claims are based on current technology. As technology improves, the mainstream consumer computers of tomorrow eventually have the processing capacity of yesterday's mainframes and suddenly the processing power required to crack the encryption becomes trivial.
As an analogy, think of encryption like a jigsaw puzzle where you have to find one specific puzzle piece. If the puzzle only has 25 pieces, it won't take you too long to accomplish. That is like a weak encryption algorithm. However, if the puzzle has 10,000 pieces it will take significantly longer.
As time goes on, though, you gather more people to join in the process and develop new strategies to sift through the pieces faster and compress the time required to look through the 10,000 pieces. That is similar to the way difficult encryption algorithms eventually become simple to crack.
There is also always the possibility of a lucky guess. The encryption cracking estimates are based on the amount of time it would take to work through every possible combination and permutation of characters to determine the encryption key. But, you could theoretically find the right key on the eighth try rather than the ten thousandth.
The fact that the A5/1 algorithm used to encrypt GSM handsets is more than two decades old and still chugging along is a testament to the strength the algorithm had at its inception. The mobile phone industry should consider itself lucky that this is only now becoming an issue.
For now, the methods revealed at the Chaos Communication Conference in Berlin still require a fairly hefty investment in technology likely to discourage any casual GSM hacking. But, the mobile phone industry as a whole needs to address the weakness of the geriatric A5/1 encryption algorithm before breaking it becomes so trivial that the encryption is completely useless.