Methods and systems for enhancing detection of fraudulent authentication data

1 . A method for enhancing security against quantum side-channel replay attacks during biometric authentication transactions comprising the steps of:

receiving, by an electronic device, data for a biometric modality of a person during an authentication transaction;

computing a feature vector from the received biometric modality data and normalizing the feature vector, the feature vector includes amplitude components;

generating a noise vector for the authentication transaction, wherein the noise vector is different for each different authentication transaction;

selecting amplitude components of the normalized feature vector to be modified;

modifying the selected amplitude components using the noise vector;

normalizing the modified feature vector;

encoding the normalized modified feature vector into qubits;

expanding, using at least one quantum algorithm, the qubits into a high-dimensional space;

detecting in the high-dimensional space anomalies indicative of fraud based on the qubits;

calculating, based on the detected anomalies, an anomaly score reflecting a likelihood that the received data is genuine;

comparing the anomaly score against a threshold value; and

in response to determining the anomaly score fails to satisfy the threshold value, determining the received biometric modality data requires secondary authentication.

2 . The method according to claim 1 , said encoding step comprises encoding the normalized modified feature vector using at least one of amplitude encoding and angle encoding.

3 . The method according to claim 1 , said selecting step comprising computing a Fisher information-based saliency metric and choosing amplitude components with the highest discriminative importance.

4 . The method according to claim 1 , said generating step comprising seeding a cryptographic pseudorandom function with a hardware-derived or a network provided nonce.

5 . The method according to claim 1 , further comprising:

determining a cryptographic state identifier calculated for each of a plurality of authentication transactions;

comparing the cryptographic state identifiers against each other; and

in response to detecting a match between any pair of cryptographic state identifiers implementing measures to minimize impacts to sensitive information.

6 . The method according to claim 1 , wherein noise-induced inaccuracies are imparted to the high-dimensional space by Noisy Intermediate-Scale Quantum hardware included in the electronic device during said expanding and detecting steps, said method further comprising reducing the noise-induced inaccuracies using at least one error mitigation technique, wherein the at least one error mitigation technique comprises zero-noise extrapolation, readout error mitigation, and randomized compiling.

7 . The method according to claim 1 , further comprising dynamically adjusting an error-mitigation parameter based on real-time metrics to maintain a target error-security trade-off.

8 . The method according to claim 1 , said encoding step comprising mapping at least sixty-four amplitude components into a corresponding qubit register.

9 . An electronic device for enhancing security against quantum side-channel replay attacks during biometric authentication transactions comprising:

a processor; and

a memory configured to store data, said electronic device being associated with a network and said memory being in communication with said processor and having instructions stored thereon which, when read and executed by said processor, cause said electronic device to:

receive data for a biometric modality of a person during an authentication transaction;

compute a feature vector from the received biometric modality data and normalize the feature vector, the feature vector includes amplitude components;

generate a noise vector for the authentication transaction, wherein the noise vector is different for each different authentication transaction;

select amplitude components of the normalized feature vector to be modified;

modify the selected amplitude components using the noise vector;

normalize the modified feature vector;

encode the normalized modified feature vector into qubits;

expand, using at least one quantum algorithm operated by the electronic device, the qubits into a high-dimensional space;

detect in the high-dimensional space anomalies indicative of fraud based on the qubits;

calculate, based on the detected anomalies, an anomaly score reflecting a likelihood that the received data is genuine;

compare the anomaly score against a threshold value; and

in response to determining the anomaly score fails to satisfy the threshold value, determine the received biometric modality data requires secondary authentication.

10 . The electronic device according to claim 9 , wherein the instructions when read and executed by said processor, cause said electronic device to encode the normalized modified feature vector using at least one of amplitude encoding and angle encoding.

11 . The electronic device according to claim 9 , wherein the instructions when read and executed by said processor, cause said electronic device to compute a Fisher information-based saliency metric and choosing amplitude components with the highest discriminative importance.

12 . The electronic device according to claim 9 , wherein the instructions when read and executed by said processor, cause said electronic device to expand the feature vectors by mapping features from the normalized modified feature vector into a high-dimensional Hilbert space.

13 . The electronic device according to claim 9 , wherein the instructions when read and executed by said processor, cause said electronic device to seed a cryptographic pseudorandom function with a hardware-derived or a network provided nonce.

14 . The electronic device according to claim 9 , wherein noise-induced inaccuracies are imparted to the high-dimensional space by Noisy Intermediate-Scale Quantum hardware included in said electronic device while expanding the normalized feature vector and detecting anomalies, and the instructions when read and executed by said processor, cause said electronic device to reduce the noise-induced inaccuracies using at least one error mitigation technique, wherein the at least one error mitigation technique comprises zero-noise extrapolation, readout error mitigation, and randomized compiling.

15 . The electronic device according to claim 9 , wherein the instructions when read and executed by said processor, cause said electronic device to:

determine a cryptographic state identifier calculated for each of a plurality of authentication transactions;

compare the cryptographic state identifiers against each other; and

in response to detecting a match between any pair of cryptographic state identifiers implement measures to minimize impacts to sensitive information.

16 . The electronic device according to claim 9 , wherein the instructions when read and executed by said processor, cause said electronic device to map at least 64 amplitude components into a corresponding qubit register.

17 . A non-transitory computer-readable recording medium in an electronic device for enhancing security against quantum side-channel replay attacks during biometric authentication transactions, the non-transitory computer-readable recording medium storing instructions which when executed by a hardware processor cause the non-transitory recording medium to perform steps comprising:

receiving data for a biometric modality of a person during an authentication transaction;

computing a feature vector from the received biometric modality data and normalizing the feature vector, the feature vector includes amplitude components;

generating a noise vector for the authentication transaction, wherein the noise vector is different for each different authentication transaction;

selecting amplitude components of the normalized feature vector to be modified;

modifying the selected amplitude components using the noise vector;

normalizing the modified feature vector;

encoding the normalized modified feature vector into qubits;

expanding, using at least one quantum algorithm, the qubits into a high-dimensional space;

detecting in the high-dimensional space anomalies indicative of fraud based on the qubits;

calculating, based on the detected anomalies, an anomaly score reflecting a likelihood that the received data is genuine;

comparing the anomaly score against a threshold value; and

in response to determining the anomaly score fails to satisfy the threshold value, determining the received biometric modality data requires secondary authentication.

18 . The non-transitory computer-readable recording medium according to claim 17 , wherein the instructions when read and executed by said processor, cause said non-transitory computer-readable recording medium to perform said encoding step by encoding the normalized modified feature vector using at least one of amplitude encoding and angle encoding.

19 . The non-transitory computer-readable recording medium according to claim 17 , wherein the instructions when read and executed by said processor, cause said non-transitory computer-readable recording medium to perform said selecting step by computing a Fisher information-based saliency metric and choosing amplitude components with the highest discriminative importance.

20 . The non-transitory computer-readable recording medium according to claim 17 , wherein the instructions when read and executed by said processor, cause said non-transitory computer-readable recording medium to perform steps comprising:

determining a cryptographic state identifier calculated for each of a plurality of authentication transactions;

comparing the cryptographic state identifiers against each other; and

in response to detecting a match between any pair of cryptographic state identifiers, implementing measures to minimize impacts to sensitive information.