A team of researchers from MIT (Massachusetts Institute of Technology) and TI (Texas Instruments) have created a new RFID (Radio-Frequency Identification) chip that they claim cannot be hacked.
The two groups say they took special care to protect against both types of attacks that plague the deployment of RFID chips found in chip-n-PIN credit cards these days.
These attacks are known by the names: side-channel attacks and glitch attacks.
Side-channel attacks occur when nearby attackers can monitor, record, or analyze data from cryptographic operations performed by the RFID chip. They can monitor for fluctuations in power consumption during these operations, as well as collect vast amounts of information, which they can later extract the cryptographic key used to secure future transactions on the chip.
The MIT and TI researchers explain that their new chip will regularly change the cryptographic key based on a random number generator, which will also run on the bank's server. The two will work in parallel, creating unique encryption keys for each card transaction.
The solution chosen by the researchers makes side-channel attacks “almost” impossible to carry out, because they would not have enough data to analyze to extract the key .
However, the researchers revealed that, despite protection against side-channel attacks, the chip could still be compromised by glitch attacks.
This type of attack works when attackers cut off the chip's power supply just before the chip wants to generate a new cryptographic key. When power is restored, the chip will use the old cryptographic key and forget about generating a new one.
In this way, attackers could repeatedly force the chip to work with the same encryption key until they gather enough information to carry out a side-channel attack.
The MIT and TI teams solved this issue by adding a 3.3-volt capacitor made of ferroelectric crystals that can store extra energy that will allow the chip to finish as many operations as possible when there is any sudden power outage and then send the data to 571 different 1.5-volt storage cells.
When power returns to the chip, it first recharges the 3.3-volt capacitor and then retrieves the previous data from the 1.5-volt cells, continuing where it left off.
After a period of testing, the MIT and TI researchers say the chips work exactly as designed. Of course, their work is only in the prototype stage right now and won't be ready for production for several years. The researchers presented their work at this year's International Solid-State Circuits Conference, held in San Francisco.
