New attacks on LTE:



Layer two (data link) does not guarantee the integrity of data. They deployed a malicious MitM relay between the mobile and the antenna to manipulate the (encrypted) payload of user data transmissions.

What kind of cryptography could be used to fix layer 2?

  • This question is a bit broad. But, generally, I guess one could apply some kind of AEAD scheme... or use an HMAC... or something like that. – hft Jul 12 '18 at 2:23
  • 4
    Like in your deleted question you reference article and paper but you did not seem to have really read it. I've even pointed you to the section Potential Countermeasure in the paper but even then you did not seem to read it - or you did not understand it. In this section the authors talk about what is needed, i.e. integrity protection. And they also point out how it can be achieved: using MAC-then-Encrypt which is already used in other places in LTE. – Steffen Ullrich Jul 12 '18 at 3:50
  • 1
    @EdS please read our How to Ask and help center pages to understand how this site works. Please stop posting multiple questions in comments, and as Steffen said, you need to read the information given. – Rory Alsop Jul 14 '18 at 20:57

As per the paper, you provided page 12 features a section titled "Potential Countermeasures" (Also mentioned by Steffen) - this section of the paper states what they require to prevent such attacks and they clearly point out how it can be done.

Even though the LTE Authentication and Key Agreement (AKA) is formally proven secure [31], this attack is still possible due to the lack of integrity protection of user plane data.

While different suitable schemes exist like AES-GCM (AES-Galois/Counter Mode), we focus on the MAC-then-Encrypt scheme that is already used for the integrity protection of the control plane. We assume that this scheme has the highest potential for being adopted in the specification - Breaking LTE on L2

MAC-then-Encrypt is a method whereby you compute the MAC (Message Authentication Code) on the plaintext, append it to the data then encrypt hence the term "MAC-then-Encrypt". If you want more details on this see: Should we MAC-then-encrypt or encrypt-then-MAC?

| improve this answer | |

Not the answer you're looking for? Browse other questions tagged or ask your own question.