Performance Analysis of Side Channel Analysis on affine masking AES implementation
Updated Implementation Plan
*********************************
Scenario 1: (Using DPA and CNN with masked variants)
*************************************************************
Step 1: Initially, we collect power traces and metadata from the ASCAD dataset(masked variants).
Step 2: Next, we extract plaintext, key, and compute SBox-based labels for the target byte from the dataset.
Step 3: Next, we normalize the traces, encode the labels, and split the dataset into training, validation, and testing sets.
Step 4: Next, we implement Differential Power Analysis (DPA) to recover key bytes for AES key byte classification.
Step 5: Next, we train CNN model using cross-entropy loss and validate it to improve accuracy.
Step 6: Next, we evaluate model performance using key rank analysis with Maximum Likelihood scoring to strengthen the key byte recovery phase.
Step 7: Finally, we plot performance for the following metrics:
7.1: Number of epochs vs. Accuracy (%)
7.2: Number of epochs vs. Precision (%)
7.3: Number of epochs vs. Recall (%)
7.4: Number of epochs vs. F1-score (%)
Scenario 2: (Using CPA and CNN with masked variants)
*************************************************************
Step 1: Initially, we collect power traces and metadata from the ASCAD dataset(masked variants).
Step 2: Next, we extract plaintext, key, and compute SBox-based labels for the target byte from the dataset.
Step 3: Next, we normalize the traces, encode the labels, and split the dataset into training, validation, and testing sets.
Step 4: Next, we implement Correlation Power Analysis (CPA) to recover key bytes for AES key byte classification.
Step 5: Next, we train CNN model using cross-entropy loss and validate it to improve accuracy.
Step 6: Next, we evaluate model performance using key rank analysis with Maximum Likelihood scoring to strengthen the key byte recovery phase.
Step 7: Finally, we plot performance for the following metrics:
7.1: Number of epochs vs. Accuracy (%)
7.2: Number of epochs vs. Precision (%)
7.3: Number of epochs vs. Recall (%)
7.4: Number of epochs vs. F1-score (%)
Scenario 3: (Using SPA and CNN with masked variants)
*************************************************************
Step 1: Initially, we collect power traces and metadata from the ASCAD dataset(masked variants).
Step 2: Next, we extract plaintext, key, and compute SBox-based labels for the target byte from the dataset.
Step 3: Next, we normalize the traces, encode the labels, and split the dataset into training, validation, and testing sets.
Step 4: Next, we implement Simple Power Analysis (SPA) to recover key bytes for AES key byte classification.
Step 5: Next, we train CNN model using cross-entropy loss and validate it to improve accuracy.
Step 6: Next, we evaluate model performance using key rank analysis with Maximum Likelihood scoring to strengthen the key byte recovery phase.
Step 7: Finally, we plot performance for the following metrics:
7.1: Number of epochs vs. Accuracy (%)
7.2: Number of epochs vs. Precision (%)
7.3: Number of epochs vs. Recall (%)
7.4: Number of epochs vs. F1-score (%)
Scenario 4: (Using DPA and CNN with Un-Masked variants)
*************************************************************
Step 1: Initially, we collect power traces and metadata from the ASCAD dataset( Unmasked variants).
Step 2: Next, we extract plaintext, key, and compute SBox-based labels for the target byte from the dataset.
Step 3: Next, we normalize the traces, encode the labels, and split the dataset into training, validation, and testing sets.
Step 4: Next, we implement Differential Power Analysis (DPA) to recover key bytes for AES key byte classification.
Step 5: Next, we train CNN model using cross-entropy loss and validate it to improve accuracy.
Step 6: Next, we evaluate model performance using key rank analysis with Maximum Likelihood scoring to strengthen the key byte recovery phase.
Step 7: Finally, we plot performance for the following metrics:
7.1: Number of epochs vs. Accuracy (%)
7.2: Number of epochs vs. Precision (%)
7.3: Number of epochs vs. Recall (%)
7.4: Number of epochs vs. F1-score (%)
Scenario 5: (Using CPA with and CNN Un-Masked variants)
*****************************************************************
Step 1: Initially, we collect power traces and metadata from the ASCAD dataset( Unmasked variants).
Step 2: Next, we extract plaintext, key, and compute SBox-based labels for the target byte from the dataset.
Step 3: Next, we normalize the traces, encode the labels, and split the dataset into training, validation, and testing sets.
Step 4: Next, we implement Correlation Power Analysis (CPA) to recover key bytes for AES key byte classification.
Step 5: Next, we train CNN model using cross-entropy loss and validate it to improve accuracy.
Step 6: Next, we evaluate model performance using key rank analysis with Maximum Likelihood scoring to strengthen the key byte recovery phase.
Step 7: Finally, we plot performance for the following metrics:
7.1: Number of epochs vs. Accuracy (%)
7.2: Number of epochs vs. Precision (%)
7.3: Number of epochs vs. Recall (%)
7.4: Number of epochs vs. F1-score (%)
Scenario 6: (Using SPA and CNN with Un-Masked variants)
*****************************************************************
Step 1: Initially, we collect power traces and metadata from the ASCAD dataset(Unmasked variants).
Step 2: Next, we extract plaintext, key, and compute SBox-based labels for the target byte from the dataset.
Step 3: Next, we normalize the traces, encode the labels, and split the dataset into training, validation, and testing sets.
Step 4: Next, we implement Simple Power Analysis (SPA) to recover key bytes for AES key byte classification.
Step 5: Next, we train CNN model using cross-entropy loss and validate it to improve accuracy.
Step 6: Next, we evaluate model performance using key rank analysis with Maximum Likelihood scoring to strengthen the key byte recovery phase.
Step 7: Finally, we plot performance for the following metrics:
7.1: Number of epochs vs. Accuracy (%)
7.2: Number of epochs vs. Precision (%)
7.3: Number of epochs vs. Recall (%)
7.4: Number of epochs vs. F1-score (%)
Scenario 7: (Using DPA and LSTM with masked variants)
*************************************************************
Step 1: Initially, we collect power traces and metadata from the ASCAD dataset(masked variants).
Step 2: Next, we extract plaintext, key, and compute SBox-based labels for the target byte from the dataset.
Step 3: Next, we normalize the traces, encode the labels, and split the dataset into training, validation, and testing sets.
Step 4: Next, we implement Differential Power Analysis (DPA) to recover key bytes for AES key byte classification.
Step 5: Next, we train the LSTM model using cross-entropy loss and validate it to improve accuracy.
Step 6: Next, we evaluate model performance using key rank analysis with Maximum Likelihood scoring to strengthen the key byte recovery phase.
Step 7: Finally, we plot performance for the following metrics:
7.1: Number of epochs vs. Accuracy (%)
7.2: Number of epochs vs. Precision (%)
7.3: Number of epochs vs. Recall (%)
7.4: Number of epochs vs. F1-score (%)
Scenario 8: (Using CPA and LSTM with masked variants)
*************************************************************
Step 1: Initially, we collect power traces and metadata from the ASCAD dataset(masked variants).
Step 2: Next, we extract plaintext, key, and compute SBox-based labels for the target byte from the dataset.
Step 3: Next, we normalize the traces, encode the labels, and split the dataset into training, validation, and testing sets.
Step 4: Next, we implement Correlation Power Analysis (CPA) to recover key bytes for AES key byte classification.
Step 5: Next, we train the LSTM model using cross-entropy loss and validate it to improve accuracy.
Step 6: Next, we evaluate model performance using key rank analysis with Maximum Likelihood scoring to strengthen the key byte recovery phase.
Step 7: Finally, we plot performance for the following metrics:
7.1: Number of epochs vs. Accuracy (%)
7.2: Number of epochs vs. Precision (%)
7.3: Number of epochs vs. Recall (%)
7.4: Number of epochs vs. F1-score (%)
Scenario 9: (Using SPA and LSTM with masked variants)
*************************************************************
Step 1: Initially, we collect power traces and metadata from the ASCAD dataset(masked variants).
Step 2: Next, we extract plaintext, key, and compute SBox-based labels for the target byte from the dataset.
Step 3: Next, we normalize the traces, encode the labels, and split the dataset into training, validation, and testing sets.
Step 4: Next, we implement Simple Power Analysis (SPA) to recover key bytes for AES key byte classification.
Step 5: Next, we train the LSTM model using cross-entropy loss and validate it to improve accuracy.
Step 6: Next, we evaluate model performance using key rank analysis with Maximum Likelihood scoring to strengthen the key byte recovery phase.
Step 7: Finally, we plot performance for the following metrics:
7.1: Number of epochs vs. Accuracy (%)
7.2: Number of epochs vs. Precision (%)
7.3: Number of epochs vs. Recall (%)
7.4: Number of epochs vs. F1-score (%)
Scenario 10: (Using DPA and LSTM with Un-Masked variants)
*******************************************************************
Step 1: Initially, we collect power traces and metadata from the ASCAD dataset( Unmasked variants).
Step 2: Next, we extract plaintext, key, and compute SBox-based labels for the target byte from the dataset.
Step 3: Next, we normalize the traces, encode the labels, and split the dataset into training, validation, and testing sets.
Step 4: Next, we implement Differential Power Analysis (DPA) to recover key bytes for AES key byte classification.
Step 5: Next, we train the LSTM model using cross-entropy loss and validate it to improve accuracy.
Step 6: Next, we evaluate model performance using key rank analysis with Maximum Likelihood scoring to strengthen the key byte recovery phase.
Step 7: Finally, we plot performance for the following metrics:
7.1: Number of epochs vs. Accuracy (%)
7.2: Number of epochs vs. Precision (%)
7.3: Number of epochs vs. Recall (%)
7.4: Number of epochs vs. F1-score (%)
Scenario 11: (Using CPA and LSTM with Un-Masked variants)
*******************************************************************
Step 1: Initially, we collect power traces and metadata from the ASCAD dataset( Unmasked variants).
Step 2: Next, we extract plaintext, key, and compute SBox-based labels for the target byte from the dataset.
Step 3: Next, we normalize the traces, encode the labels, and split the dataset into training, validation, and testing sets.
Step 4: Next, we implement Correlation Power Analysis (CPA) to recover key bytes for AES key byte classification.
Step 5: Next, we train the LSTM model using cross-entropy loss and validate it to improve accuracy.
Step 6: Next, we evaluate model performance using key rank analysis with Maximum Likelihood scoring to strengthen the key byte recovery phase.
Step 7: Finally, we plot performance for the following metrics:
7.1: Number of epochs vs. Accuracy (%)
7.2: Number of epochs vs. Precision (%)
7.3: Number of epochs vs. Recall (%)
7.4: Number of epochs vs. F1-score (%)
Scenario 12: (Using SPA and LSTM with Un-Masked variants)
*******************************************************************
Step 1: Initially, we collect power traces and metadata from the ASCAD dataset(Unmasked variants).
Step 2: Next, we extract plaintext, key, and compute SBox-based labels for the target byte from the dataset.
Step 3: Next, we normalize the traces, encode the labels, and split the dataset into training, validation, and testing sets.
Step 4: Next, we implement Simple Power Analysis (SPA) to recover key bytes for AES key byte classification.
Step 5: Next, we train the LSTM model using cross-entropy loss and validate it to improve accuracy.
Step 6: Next, we evaluate model performance using key rank analysis with Maximum Likelihood scoring to strengthen the key byte recovery phase.
Step 7: Finally, we plot performance for the following metrics:
7.1: Number of epochs vs. Accuracy (%)
7.2: Number of epochs vs. Precision (%)
7.3: Number of epochs vs. Recall (%)
7.4: Number of epochs vs. F1-score (%)
Software requirement:
*************************
1. Development Tool: Python 3.11.4 or above
2. Operating System: Windows 10 (64-bit) or above
Dataset Link:
***************
(Masked) :For Scenario 1,2,3,7,8,9
https://static.data.gouv.fr/resources/ascad-atmega-8515-variable-key/20190903-083349/ascad-variable.h5
(Unmasked) : For Scenario 4,5,6,10,11,12
https://static.data.gouv.fr/resources/ascad-atmega-8515-variable-key/20190903-084119/ascad-variable-desync50.h5
Note :-
*******
1) If the plan does not meet your requirements, provide detailed steps, parameters, models, or expected results in advance. Once implemented, changes won’t be possible without prior input; otherwise, we’ll proceed as per our implementation plan.
2) If the plan satisfies your requirement, Please confirm with us.
3) Project based on Simulation only, not a real time project.
4) If you have any changes in the Dataset , kindly provide before implementation.Our work is completely based on dataset values.
5) Please understand that any modifications made to the confirmed implementation plan will not be made after the project development.
