//This is a version to match Ha's noise, not sure if right
public static void GenerateReliabilityTrainingDataHaWay(XORArbiterPUF xPUF, int numberOfMeasurements, double[][] trainingData, double[][] trainingReliability, Random randomGenerator)
{
int trainingSize = trainingData.Length;
int bitNum = xPUF.BitNumber;
//pregenerate training data inputs
double[] noisyAPUFWeight1 = new double[bitNum + 1];
double[] noisyAPUFWeight2 = new double[bitNum + 1];
double[] sumOfResponses = new double[trainingSize];
for (int i = 0; i < trainingSize; i++)
{
trainingData[i] = GenerateRandomPhiVector(bitNum, randomGenerator);
}
for (int m = 0; m < numberOfMeasurements; m++) //Ha's way is flipped
{
double[] noiseAPUF1 = new double[bitNum + 1];
double[] noiseAPUF2 = new double[bitNum + 1];
for (int i = 0; i < noiseAPUF1.Length; i++)
{
noiseAPUF1[i] = GenerateRandomNormalVariable(0, 0.1, randomGenerator);
noiseAPUF2[i] = GenerateRandomNormalVariable(0, 0.1, randomGenerator);
}
double[] originalAPUFWeight1 = xPUF.GetAllGroundTruthWeights()[0];
double[] originalAPUFWeight2 = xPUF.GetAllGroundTruthWeights()[1];
//Combine the noisy and original weights
for (int i = 0; i < noisyAPUFWeight1.Length; i++)
{
noisyAPUFWeight1[i] = originalAPUFWeight1[i] + noiseAPUF1[i];
noisyAPUFWeight2[i] = originalAPUFWeight2[i] + noiseAPUF2[i];
}
ArbiterPUF aNoisy1 = new ArbiterPUF(noisyAPUFWeight1);
ArbiterPUF aNoisy2 = new ArbiterPUF(noisyAPUFWeight2);
//Compute for each sample
for (int i = 0; i < trainingSize; i++)
{
int cc = aNoisy1.ComputeResponse(trainingData[i]);
int ccc = aNoisy2.ComputeResponse(trainingData[i]);
int result = aNoisy1.ComputeResponse(trainingData[i]) ^ aNoisy2.ComputeResponse(trainingData[i]);
sumOfResponses[i] = sumOfResponses[i] + result;
}
}
//Last compute the reliability
for (int i = 0; i < trainingSize; i++)
{
trainingReliability[i] = new double[1];
//trainingReliability[i][0]= Math.Abs(numberOfMeasurements / 2.0 - (sumOfResponses[i] / (double)numberOfMeasurements));
trainingReliability[i][0] = Math.Abs(numberOfMeasurements / 2.0 - (sumOfResponses[i]));
}
//for (int i = 0; i < trainingSize; i++)
//{
// double sumOfResponses = 0;
// //trainingData[i] = GenerateRandomPhiVector(bitNum);
// trainingData[i] = GenerateRandomPhiVector(bitNum, randomGenerator);
// for (int m = 0; m < numberOfMeasurements; m++)
// {
// double randomNoise
// sumOfResponses = sumOfResponses + aPUF.ComputeNoisyResponse(trainingData[i]); //sum the measurements
// }
// trainingReliability[i] = new double[1];
// //trainingReliability[i][0] = sumOfResponses / (double)numberOfMeasurements;
// trainingReliability[i][0] = Math.Abs(numberOfMeasurements / 2.0 - (sumOfResponses / (double)numberOfMeasurements));
//}
}
