0
\$\begingroup\$

Some Background:

I am working on an embedded c project for a samd21j18a MCU. The MCU communicates with 2 crypto eeproms on the i2c bus. All the data is encrypted with AES in CCM and additional authentication uses SHA256. It is safe to say that if the bus is sniffed, the data is secure. My question is that during some authentication checking it is possible that EEPROM can return a true/false value for authentication. This is an optional means of authenticating. It is possible that a device can sniff a bus and also spoof a valid response?

\$\endgroup\$

1 Answer 1

2
\$\begingroup\$

Of course it's possible, at worst by cutting the wires to the original peripheral and interposing something with a controllable pass-gate, or perhaps even just a strong push-pull driver able to overpower the original peripheral.

This (perhaps without disconnecting the original) is for example what the "Bus Pirate" series of debug boards was designed to do.

It would also be very possible to make something with some other MCU controlling pass gates, or a small FPGA alternately passing through or overriding the peripheral side of the bus.

Unless your messages have end-to-end encryption and cryptographic validation between chips you have to consider that they are intercept-able and spoof-able. If someone is sufficiently determined and budgeted, even on-chip signals could be. And without a lot of care, the behavior of the peripheral chip could turn out to be fully emulatable.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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