Relay attack prevention

1. A system comprising:

a hardware processor; and

a non-transitory computer-readable storage medium having stored thereon program instructions, the program instructions executable by the at least one hardware processor to:

receive, using a first transceiver, a radio-frequency (RF) transmission from an RF device; and

detect a second transceiver performing a relay or man-in-the-middle (MITM) attack by:

detecting double-Automatic Gain Control (AGC) distortion effect in the received RF transmission by:

finding and extracting a rise transient from a beginning to a start of a data region of the received RF transmission;

estimating a differential AGC distortion caused by the receiver, based on the known output of an AGC circuit if the system; and

determining whether the received RF transmission exhibits greater AGC distortion than the differential AGC distortion such as to indicate distortion from a relay or replay AGC stage of the second transceiver.

2. The system of claim 1 , wherein estimating the differential AGC distortion caused by the receiver and determining whether the received RF transmission exhibits greater AGC distortion than the differential AGC distortion includes measuring mean power of complex column sample vector.

3. The system of claim 1 , wherein detecting double-AGC distortion further comprises:

finding and extracting samples of the received RF transmission including:

before the beginning of the rise transient;

after a fall transient; and

from frequencies out of a bandwidth of the received RF transmission during a data transmission section of the received RF transmission.

4. The system of claim 3 , wherein the samples are extracted by using a high pass filter with a cutoff frequency that is higher than the bandwidth of the received RF transmission.

5. The system of claim 1 , wherein detecting double-AGC distortion further comprises:

determining that the received RF transmission was not relayed or replayed by determining that the received RF transmission was not previously distorted by a AGC stage of another RF device prior to the first RF receiver receiving the RF transmission.

6. The system of claim 1 , wherein detecting double-AGC distortion further comprises:

training one or more machine learning classifiers for detecting double-AGC distortion with a set of clean signals, which are signals that not distorted by an AGC circuit.

7. The system of claim 1 , wherein detecting double-AGC distortion further comprises at least one of:

determining a noise mean power;

assuming an underlying Rayleigh model and performing a parameter estimation by a Maximum Likelihood or a Method of Moments approach; or

assuming a real normal model and calculating variances, an average and squared root.

8. A method comprising operating at least one hardware processor for:

receiving, using a first transceiver, a radio-frequency (RF) transmission from an RF device; and

detecting a second transceiver performing a relay or man-in-the-middle (MITM) attack by:

detecting double-Automatic Gain Control (AGC) distortion effect in the received RF transmission by:

finding and extracting a rise transient from a beginning to a start of a data region of the received RF transmission;

estimating a differential AGC distortion caused by the receiver, based on the known output of an AGC circuit if the system; and

determining whether the received RF transmission exhibits greater AGC distortion than the differential AGC distortion such as to indicate distortion from a relay or replay AGC stage of the second transceiver.

9. The method of claim 8 , wherein estimating the differential AGC distortion caused by the receiver and determining whether the received RF transmission exhibits greater AGC distortion than the differential AGC distortion includes measuring mean power of complex column sample vector.

10. The method of claim 8 , wherein detecting double-AGC distortion further comprises:

finding and extracting samples of the received RF transmission including:

before the beginning of the rise transient;

after a fall transient; and

from frequencies out of a bandwidth of the received RF transmission during a data transmission section of the received RF transmission.

11. The method of claim 10 , wherein the samples are extracted by using a high pass filter with a cutoff frequency that is higher than the bandwidth of the received RF transmission.

12. The method of claim 8 , wherein detecting double-AGC distortion further comprises:

determining that the received RF transmission was not relayed or replayed by determining that the received RF transmission was not previously distorted by a AGC stage of another RF device prior to the first RF receiver receiving the RF transmission.

13. The method of claim 8 , wherein the detecting double-AGC distortion further comprises:

training one or more machine learning classifiers for detecting double-AGC distortion with a set of clean signals, which are signals that not distorted by an AGC circuit.

14. The method of claim 8 , wherein

detecting double-AGC distortion further comprises at least one of:

determining a noise mean power;

assuming an underlying Rayleigh model and performing a parameter estimation by a Maximum Likelihood or a Method of Moments approach; or

assuming a real normal model and calculating variances, an average and squared root.

15. A computer program product comprising a non-transitory computer-readable storage medium having program instructions embodied therewith, the program instructions executable by at least one hardware processor to:

receive, using a first transceiver, a radio-frequency (RF) transmission from an RF device; and

detect a second transceiver performing a relay or man-in-the-middle (MITM) attack by:

detecting double-Automatic Gain Control (AGC) distortion effect in the received RF transmission by:

finding and extracting a rise transient from a beginning to a start of a data region of the received RF transmission;

estimating a differential AGC distortion caused by the receiver, based on the known output of an AGC circuit if the system; and

determining whether the received RF transmission exhibits greater AGC distortion than the differential AGC distortion such as to indicate distortion from a relay or replay AGC stage of the second transceiver.

16. The computer program product of claim 15 , wherein estimating the differential AGC distortion caused by the receiver and determining whether the received RF transmission exhibits greater AGC distortion than the differential AGC distortion includes measuring mean power of complex column sample vector.

17. The computer program product of claim 15 , wherein detecting double-AGC distortion further comprises:

finding and extracting samples of the received RF transmission including:

before the beginning of the rise transient;

after a fall transient; and

from frequencies out of a bandwidth of the received RF transmission during a data transmission section of the received RF transmission,

wherein the samples are extracted by using a high pass filter with a cutoff frequency that is higher than the bandwidth of the received RF transmission.

18. The computer program product of claim 15 , wherein detecting double-AGC distortion further comprises:

determining that the received RF transmission was not relayed or replayed by determining that the received RF transmission was not previously distorted by a AGC stage of another RF device prior to the first RF receiver receiving the RF transmission.

19. The computer program product of claim 15 , wherein detecting double-AGC distortion further comprises:

training one or more machine learning classifiers for detecting double-AGC distortion with a set of clean signals, which are signals that not distorted by an AGC circuit.

20. The computer program product of claim 15 , wherein detecting double-AGC distortion further comprises:

determining a noise mean power;

assuming an underlying Rayleigh model and performing a parameter estimation by a Maximum Likelihood or a Method of Moments approach; or

assuming a real normal model and calculating variances, an average and squared root.