JP5397625B2 - Side channel attack resistance evaluation apparatus, method and program thereof - Google Patents

Description

  The present invention relates to a side channel attack resistance evaluation apparatus, a method thereof, and a program thereof that perform resistance evaluation against a side channel attack based on side channel information leaked from a cryptographic apparatus.

  As information is converted into electronic data, encryption is an indispensable technology for information protection and confidential communication. In order to maintain the security of the cipher, it is necessary to prevent secret information such as a key from being easily guessed. Here, methods such as full key search, linear cryptanalysis that performs mathematical cryptanalysis, and differential cryptanalysis are known as cryptanalysis methods, but these methods cannot be analyzed in real time. Is the situation.

  On the other hand, on the assumption that an attacker can accurately measure side channel information such as processing time and power consumption in an IC card with encryption function or a device such as a portable terminal that implements encryption, the side channel Side-channel attacks that attempt to acquire confidential information from information and countermeasures have become a major research theme.

  The side channel information includes information related to the processing and data being executed in the cryptographic device that is the target of attack. By analyzing the side channel information, internal processing of encryption and confidential information are analyzed. Is possible. For example, the encryption algorithm, processing timing, and secret key can be estimated.

  Specific attack methods of the side channel attack include timing attacks focusing on processing time (see Non-Patent Document 1), power analysis focusing on power consumption (see Non-Patent Document 2), and electromagnetic wave analysis focusing on leakage electromagnetic waves. (Refer nonpatent literature 3) etc. are known.

  As described above, since side channel attacks are widely known, an apparatus equipped with a cipher requires practically resistance to attack against side channel attacks (hereinafter referred to as “tamper resistance”). For this reason, research on tamper resistance technology that makes it difficult to estimate secret information such as encryption algorithms from side channel information is underway.

Here, tamper resistance refers to the performance of preventing leakage of confidential information and modification of functions against attacks.
JP 2007-116215 A P. Kocher “Timing Attacks on Implementations of Diffie-Hellman, RSA, DSS, and Other Systems,” Crypto'96, pp.101-113, 1996. P. Kocher, J. Jaffe and B. Jun, “Introduction to Differential Power Analysis and Related Attacks,” 1998. Karine Gandolfi, Christophe Mourtel, and Francis Olivier, “Electromagnetic Analysis: Concrete Results,” In the Proceedings of the Workshop on Cryptographic Hardware and Embedded Systems 2001 (CHES 2001), LNCS 2162 Paris, France, May 2001, pp 251-261. Hideki Imai “Information Security-For a Safe and Secure Society”, IEICE Journal, Vol.90, No.5, pp.334-339

  In this regard, as a tamper-resistant technique, a tamper-resistant method (see Patent Document 1) that prevents leakage of confidential information from side channel information by arbitrarily adding unnecessary information to the side channel information has been proposed. Here, it is necessary to evaluate the effectiveness of the tamper-resistant technique as to whether the resistance to side channel attacks is actually improved by applying the tamper-resistant technique as described above.

  In addition, in order to confirm whether or not the installed cryptographic device is really safe, a technique for evaluating resistance to side channel attacks is required (see Non-Patent Document 4).

  However, when evaluating side channel attack tolerance, unnecessary information such as measurement noise and environmental noise may be added to the side channel information, so that characteristics indicating confidential information used for resistance evaluation may not be extracted from the side channel information. . Therefore, in order to accurately evaluate the resistance to side channel attacks, it is necessary to remove unnecessary information included in the side channel information.

  Accordingly, an object of the present invention is to provide a side channel attack resistance evaluation device capable of suppressing deterioration in accuracy of resistance evaluation due to unnecessary information on side channel information and accurately evaluating whether or not a side channel attack can be performed on a cryptographic device. And

According to the first aspect of the present invention, in a side channel attack resistance evaluation apparatus that performs resistance evaluation against a side channel attack performed using side channel information leaked from an encryption apparatus as an apparatus, the evaluation target encryption Side channel information measuring means for measuring side channel information generated from the apparatus, and side channel information obtained by performing signal processing for noise reduction on the side channel information obtained by the side channel information measuring means, A side that determines whether or not the cryptographic device to be evaluated is resistant to side channel attacks based on the signal processing means that outputs the signal, side-channel information that has been processed by the signal processing means, and a reference waveform A channel attack resistance evaluation means, and a side channel attack resistance evaluation characterized by comprising: Location is provided.

According to a second aspect of the present invention, there is provided a side channel attack resistance evaluation method for evaluating resistance to a side channel attack performed using side channel information leaked from a cryptographic device as a method. Side channel information measurement step for measuring side channel information generated from the device, and side channel information obtained by performing signal processing for noise reduction on the side channel information obtained in the side channel information measurement step. A side for determining whether or not the cryptographic device to be evaluated is resistant to side channel attacks based on the signal processing step for outputting the signal, the side channel information on which the signal processing has been performed output in the signal processing step, and the reference waveform A channel attack resistance judging step, Yaneru attack resistance evaluation method is provided.

According to a third aspect of the present invention, in a side channel attack resistance evaluation program for evaluating resistance to a side channel attack performed using side channel information leaked from an encryption device as a program, Side channel information measurement function for measuring side channel information generated from a device, and side channel information obtained by performing signal processing for noise reduction on the side channel information obtained by the side channel information measurement function The side for determining whether or not the cryptographic device to be evaluated is resistant to side channel attacks based on the signal processing function for outputting the signal, the side channel information after the signal processing output in the signal processing function, and the reference waveform It is characterized by having a computer implement a channel attack resistance determination function. Side channel attack tolerance evaluation program is provided that.

  According to the present invention, by removing unnecessary information such as noise included in the side channel information, it becomes easy to extract features used for evaluation from the side channel information, and the evaluation accuracy of the evaluation target side channel attack resistance is improved. It becomes possible to improve.

It is a schematic block diagram of the side channel attack tolerance evaluation apparatus which can apply this invention. It is the flowchart which showed the side channel attack tolerance determination means which concerns on the 1st Embodiment of this invention. It is the flowchart which showed the side channel attack tolerance determination means which concerns on the 2nd Embodiment of this invention. It is the flowchart which showed the side channel attack tolerance determination means which concerns on the 3rd Embodiment of this invention. It is the flowchart which showed the side channel attack tolerance determination means which concerns on the 4th Embodiment of this invention. It is the flowchart which showed the side channel attack tolerance determination means which concerns on the 5th Embodiment of this invention. It is the flowchart which showed the side channel attack tolerance determination means which concerns on the 6th Embodiment of this invention. It is the flowchart which showed the side channel attack tolerance determination means which concerns on the 7th Embodiment of this invention. It is the flowchart which showed the side channel attack tolerance determination means based on the 8th Embodiment of this invention. It is the flowchart which showed the side channel attack tolerance determination means based on the 9th Embodiment of this invention. It is the flowchart which showed the side channel attack tolerance determination means based on the 10th Embodiment of this invention. It is the flowchart which showed the side channel attack tolerance determination means based on the 11th Embodiment of this invention. It is the flowchart which showed the side channel attack tolerance determination means based on the 12th Embodiment of this invention. It is the flowchart which showed the side channel attack tolerance determination means based on the 13th Embodiment of this invention. In Example 1, it is a figure showing the reference | standard waveform used with the side channel attack tolerance evaluation apparatus. It is a figure showing the electromagnetic wave measurement waveform at the time of AES encryption processing. It is a figure showing the power spectrum waveform in the electromagnetic wave measurement waveform at the time of AES encryption processing. It is a figure showing the band pass filter application waveform with respect to the electromagnetic wave measurement waveform at the time of AES encryption processing. It is a figure showing the correlation calculation result of the bandpass filter application waveform with respect to the electromagnetic wave measurement waveform at the time of AES encryption processing, and a reference | standard waveform. It is a figure showing the average waveform of an electric power waveform when A5 / 1 encryption processing is measured 100 times. It is a figure showing the median filter application waveform of the window width 101 with respect to the average waveform of an electric power waveform when A5 / 1 encryption processing is measured 100 times. It is a figure showing the power measurement waveform at the time of A5 / 1 encryption processing. It is a figure showing the median filter application waveform of the window width 3 with respect to the electric power measurement waveform at the time of A5 / 1 encryption processing. It is a figure showing the correlation calculation result of the electric power measurement waveform (median filter application of window width 3) at the time of A5 / 1 encryption processing, and a reference waveform. It is a figure showing the change of a correlation value when changing the window width of a median filter in the correlation with the electric power measurement waveform at the time of A5 / 1 encryption processing to which a median filter is applied, and a reference waveform. It is a figure showing the median filter application waveform of the window width 55 with respect to the electric power measurement waveform at the time of A5 / 1 encryption processing. It is a figure showing the correlation calculation result of the electric power measurement waveform (median filter application of the window width 55) at the time of A5 / 1 encryption processing, and a reference | standard waveform. It is a figure showing the correlation calculation result of the electric power measurement waveform at the time of A5 / 1 encryption processing (median filter non-application) and a reference waveform.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Encryption apparatus 2 Side channel information measuring apparatus 3 Signal processing part 4 Side channel attack tolerance evaluation part

  Next, the best mode for carrying out the present invention will be described in detail with reference to the drawings.

  FIG. 1 is a diagram showing a schematic configuration of a side channel attack resistance evaluation apparatus to which the present invention can be applied.

  Referring to FIG. 1, the side channel attack resistance evaluation device includes an evaluation target encryption device 1, a side channel information measurement device 2, a signal processing unit 3, and a side channel attack resistance evaluation unit 4.

  The encryption apparatus 1 is an apparatus that performs encryption / decryption processing such as encryption for plain text and decryption for cipher text. As the encryption device 1, various information processing devices that execute encryption / decryption processing can be employed. For example, there are a personal computer (PC), a portable terminal, a contact type and non-contact type IC card, a reader / writer, and the like.

  The side channel information measuring device 2 is a device that measures side channel information leaked when the encryption device 1 performs encryption / decryption processing.

  As the side channel information, various kinds of information affected by internal processing in the encryption apparatus can be adopted. For example, power, electromagnetic waves, sound, temperature, etc. can be exemplified. If electromagnetic waves are used as side channel information, an oscilloscope, spectrum analyzer, or the like can be used as the side channel information measuring device 2.

  The signal processing unit 3 performs signal processing on the side channel information input from the side channel information measuring apparatus 2 based on the parameters.

  Examples of signal processing include signal processing using a filter based on an average value, a filter based on a median, and a frequency filter. Examples of the signal processing include signal processing using a low-pass filter, a high-pass filter, a band-pass filter, a notch filter, a multi-band filter, a filter based on a wavelet transform, and the like.

  The signal processing parameters include, for example, (1) the signal processing method to be applied, (2) the window width and load of the filter, the number of stages, (3) the pass / stop band in the frequency domain, and (4) the wavelet level. , Base, threshold, etc.

  These parameters are set in advance by the evaluator or in advance in the side channel attack resistance evaluation apparatus. In addition, the signal processing unit 3 may have a function of changing the parameter according to “side channel information input from the side channel information measuring device 2” or “determination result of the side channel attack resistance evaluation unit 4”. Is possible.

  The side channel attack resistance evaluation unit 4 performs resistance evaluation against a side channel attack on the signal-processed side channel information input from the signal processing unit 3.

  As an evaluation method, a side channel attack is performed on signal-processed side channel information, and the result of determining whether or not the attack is successful, or a method of using correlation, distance, and similarity with a reference waveform Etc. Examples of the side channel attack method include timing attack, electromagnetic wave analysis, simple power analysis, differential power analysis, failure use attack, cache attack, acoustic analysis attack, and the like.

  In addition, when carrying out tolerance evaluation, when a plurality of side channel information is required, the measurement by the side channel information measurement device 2 and the processing by the signal processing unit 3 are performed a plurality of times, thereby Channel information can be acquired. Further, when the signal processing unit 3 has a function of changing a parameter according to the content of resistance evaluation, the side channel attack resistance evaluation unit 4 includes the content of resistance evaluation of the side channel attack (presence / absence of resistance, correlation value). , Distance, similarity, etc.) to the signal processing unit 3.

  Next, the operation of the side channel attack resistance evaluation apparatus configured as described above will be described.

  First, parameters used in the signal processing unit 3 are set. Further, the evaluation method used by the side channel attack resistance evaluation unit 4 is determined.

  Thereafter, encryption processing is executed by the encryption device 1 to be evaluated, and side channel information leaked from the encryption device 1 is measured by the side channel information measurement device 2. The measured side channel information is processed by the signal processing unit 3, and the signal processed side channel information from which unnecessary information is removed is output to the side channel attack resistance evaluation unit 4.

  Finally, the side channel attack resistance evaluation unit 4 performs the resistance evaluation of the encryption device 1 against the side channel attack using the signal-processed side channel information.

[First Embodiment]
In the first embodiment of the present invention to be described subsequently, the signal processing unit 3 performs signal processing on the side channel information measured by the encryption device 1 to be evaluated using parameters determined in advance. To do. Also, the side channel attack resistance evaluation unit 4 performs side channel attack resistance evaluation of the cryptographic device 1 using the signal-processed side channel information.

  FIG. 2 is a flowchart of the side channel attack resistance evaluation means in Embodiment 1 of the present invention.

  First, when the determination process is started (step A1), side channel information is input to the side channel information measuring apparatus 2 (step A2).

  When the input of the side channel information is completed, the signal processing unit 3 performs signal processing on the input side channel information based on a predetermined parameter (step A3).

  When the signal processing is completed, the side channel attack resistance evaluation unit 4 evaluates the side channel attack resistance using the signal processed side channel information (step A4).

  As a result of the evaluation, if it is determined that there is resistance, the cryptographic apparatus 1 is determined to have side channel attack resistance and the process ends (steps A5 and A6).

  On the other hand, if it is determined that there is no resistance as a result of the evaluation, it is determined that the encryption apparatus 1 has no side channel attack resistance and the process ends (steps A7 and A8).

  By implementing the first embodiment described above, even when unnecessary information is added to the side channel information, unnecessary information is removed by performing signal processing. It is easy to extract from the side channel information. Therefore, the evaluation accuracy of side channel attack resistance is improved. Here, the evaluation accuracy includes, for example, the correct answer rate of the determination result, the number of data necessary to obtain a certain correct answer rate, and the like. And if the correct answer rate improves or the number of necessary data decreases, it can be said that the evaluation accuracy has improved. That is, by using the embodiment of the present invention, even if the analysis is performed based on the same number of data, there is an effect that the correct answer rate is improved as compared with the case where the present embodiment is not used. In addition, the number of measurement data necessary for obtaining the same correct answer rate can be reduced.

  Further, when the tamper resistant method is applied to the encryption device 1, it is possible to evaluate whether or not the applied tamper resistant method is effective in the cryptographic device 1. If unnecessary information is removed in the signal processing unit 3 and the side channel attack resistance evaluation unit 4 determines that there is no resistance, it can be said that the effectiveness of the tamper resistance method is low. On the other hand, even if the signal processing unit 3 performs processing, if the side channel attack resistance evaluation unit 4 determines that there is resistance, the tamper resistance method is highly effective.

[Second Embodiment]
In the second embodiment of the present invention to be described subsequently, the signal processing unit 3 applies a filter based on the average value of the side channel information to the side channel information measured by the encryption device 1 to be evaluated. Thereafter, the side channel attack resistance evaluation unit 4 performs side channel attack resistance evaluation of the cryptographic device 1 using the side channel information subjected to signal processing.

  FIG. 3 is a flowchart of the side channel attack resistance evaluation means in Embodiment 2 of the present invention.

  The difference from the first embodiment is that a filter based on an average value is applied after the side channel information is input in step A2 (step A3-1).

  By applying a filter based on the average value to the side channel information, the side channel information is smoothed, and unnecessary information such as noise can be suppressed. In particular, when unnecessary information is Gaussian noise, this embodiment in which a filter based on an average value is applied is effective. By suppressing unnecessary information, the side channel attack resistance evaluation unit 4 can perform accurate resistance evaluation.

[Third Embodiment]
In the third embodiment of the present invention to be described subsequently, the signal processing unit 3 applies a filter based on the median of the side channel information to the side channel information measured by the encryption device 1 to be evaluated. Thereafter, the side channel attack resistance evaluation unit 4 performs side channel attack resistance evaluation of the cryptographic device 1 using the side channel information subjected to signal processing.

  FIG. 4 is a flowchart of the side channel attack resistance evaluation means in Embodiment 3 of the present invention.

  A difference from the first embodiment is that a filter based on the median is applied after the input of the side channel information in Step A2 (Step A3-2).

  By applying a median filter to the side channel information, unnecessary information such as noise included in the side channel information can be suppressed. In particular, when unnecessary information is impulsive noise, this embodiment in which a filter based on median is applied is effective. By suppressing unnecessary information, the side channel attack resistance evaluation unit 4 can perform accurate resistance evaluation.

[Fourth Embodiment]
In the fourth embodiment of the present invention to be described subsequently, the signal processing unit 3 applies a frequency filter to the side channel information measured by the encryption device 1 to be evaluated. Thereafter, the side channel attack resistance evaluation unit 4 performs side channel attack resistance evaluation of the cryptographic device 1 using the side channel information subjected to signal processing.

  FIG. 5 is a flowchart of the side channel attack resistance evaluation means in Embodiment 4 of the present invention.

  The difference from the first embodiment is that a frequency filter is applied after the input of the side channel information in step A2 (step A3-3).

  While applying the frequency filter to the side channel information to attenuate the frequency band of unnecessary information included in the side channel information, while passing the necessary information band, while saving the necessary information on the side channel information, It is possible to suppress unnecessary information. By suppressing unnecessary information, the side channel attack resistance evaluation unit 4 can perform accurate resistance evaluation.

[Fifth Embodiment]
In the fifth embodiment of the present invention to be described subsequently, the signal processing unit 3 applies a low-pass filter to the side channel information measured by the encryption device 1 to be evaluated. Thereafter, the side channel attack resistance evaluation unit 4 performs side channel attack resistance evaluation of the cryptographic device 1 using the side channel information subjected to signal processing.

  FIG. 6 is a flowchart of the side channel attack resistance evaluation means in Embodiment 5 of the present invention.

  The difference from the fourth embodiment is that a low-pass filter is applied after the input of the side channel information in step A2 (step A3-3a).

  By applying a low pass filter to the side channel information to attenuate the frequency band of unnecessary information included in the side channel information, it is possible to suppress unnecessary information included in the side channel information. In particular, when unnecessary information is added to the high frequency band, the present embodiment in which the low pass filter is applied is effective. By suppressing unnecessary information, the side channel attack resistance evaluation unit 4 can perform accurate resistance evaluation.

[Sixth Embodiment]
In the sixth embodiment of the present invention to be described subsequently, the signal processing unit 3 applies a high-pass filter to the side channel information measured by the encryption device 1 to be evaluated. Thereafter, the side channel attack resistance evaluation unit 4 performs side channel attack resistance evaluation of the cryptographic device 1 using the side channel information subjected to signal processing.

  FIG. 7 is a flowchart of the side channel attack resistance evaluation means in Embodiment 6 of the present invention.

  The difference from the fourth embodiment is that a high-pass filter is applied after the input of side channel information in step A2 (step A3-3b).

  By applying a high-pass filter to the side channel information to attenuate the frequency band of unnecessary information included in the side channel information, it is possible to suppress unnecessary information included in the side channel information. In particular, when the features necessary for evaluation appear as impulses, the necessary information exists in the high frequency band. Therefore, by applying the high-pass filter, unnecessary low frequency band components are suppressed, and the side channel attack resistance evaluation unit 4 can perform accurate resistance evaluation.

[Seventh Embodiment]
In the seventh embodiment of the present invention to be described subsequently, the signal processing unit 3 applies a bandpass filter to the side channel information measured by the encryption device 1 to be evaluated. Thereafter, the side channel attack resistance evaluation unit 4 performs side channel attack resistance evaluation of the cryptographic device 1 using the side channel information subjected to signal processing.

  FIG. 8 is a flowchart of the side channel attack resistance evaluation means in Embodiment 7 of the present invention.

  The difference from the fourth embodiment is that a band pass filter is applied after the side channel information is input in step A2 (step A3-3c).

  By applying a band pass filter to the side channel information and passing only necessary information on the side channel information, unnecessary information included in the side channel information can be suppressed. In particular, when a characteristic necessary for evaluation is expressed by a frequency in a certain band, by passing only that band, unnecessary information is suppressed, and the side channel attack resistance evaluation unit 4 can perform an accurate resistance evaluation. .

[Eighth Embodiment]
In the eighth embodiment of the present invention to be described subsequently, the signal processing unit 3 applies a notch filter to the side channel information measured by the encryption device 1 to be evaluated. Thereafter, the side channel attack resistance evaluation unit 4 performs side channel attack resistance evaluation of the cryptographic device 1 using the side channel information subjected to signal processing.

  FIG. 9 is a flowchart of the side channel attack resistance evaluation means in Embodiment 8 of the present invention.

  The difference from the fourth embodiment is that a notch filter is applied after the input of the side channel information in step A2 (step A3-3d).

  By applying a notch filter to the side channel information and attenuating the frequency band of unnecessary information included in the side channel information, it is possible to suppress unnecessary information included in the side channel information. In particular, when unnecessary information is concentrated in a specific band, by attenuating only that band, unnecessary information is suppressed, and the side channel attack resistance evaluation unit 4 can perform accurate resistance evaluation.

[Ninth Embodiment]
In the ninth embodiment of the present invention to be described subsequently, the signal processing unit 3 applies a multiband filter to the side channel information measured by the encryption device 1 to be evaluated. Thereafter, the side channel attack resistance evaluation unit 4 performs side channel attack resistance evaluation of the cryptographic device 1 using the side channel information subjected to signal processing.

  FIG. 10 is a flowchart of the side channel attack resistance evaluation means in Embodiment 9 of the present invention.

  The difference from the fourth embodiment is that a multiband filter is applied after the input of the side channel information in step A2 (step A3-3e).

  By applying a multiband filter to the side channel information to attenuate the frequency band of unnecessary information included in the side channel information, it is possible to suppress unnecessary information included in the side channel information. In particular, when necessary information is concentrated in a plurality of specific bands, the necessary information is saved while unnecessary information is suppressed. Therefore, the side channel attack resistance evaluation unit 4 can perform accurate resistance evaluation. It becomes.

[Tenth embodiment]
In the tenth embodiment of the present invention to be described subsequently, the signal processing unit 3 applies a filter based on wavelet transform to the side channel information measured by the evaluation target encryption device 1. Thereafter, the side channel attack resistance evaluation unit 4 performs side channel attack resistance evaluation of the cryptographic device 1 using the side channel information subjected to signal processing.

  FIG. 11 is a flowchart of the side channel attack resistance evaluation means in Embodiment 10 of the present invention.

  The difference from the first embodiment is that a filter based on wavelet transform is applied after the input of the side channel information in step A2 (step A3-4).

  In the signal processing unit 3, first, side channel information is recursively divided into a high frequency component and a low frequency component by wavelet transform to generate transform coefficients of a plurality of band components, and then processing is performed on the transform coefficients carry out. For example, a coefficient that falls below a predetermined threshold is suppressed with respect to a conversion coefficient existing in a high frequency component. Then, the side channel information is reconstructed using the processed coefficients. Since unnecessary information is suppressed by performing wavelet transform and performing processing on the transform coefficient, the side channel attack resistance evaluation unit 4 can perform accurate resistance evaluation.

[Eleventh embodiment]
In the eleventh embodiment of the present invention to be described subsequently, the signal processing unit 3 performs feature discrimination on the side channel information measured by the encryption device 1 to be evaluated. Then, a signal processing parameter to be used is determined according to the determined feature, and the signal processing is performed on the side channel information. Thereafter, the side channel attack resistance evaluation unit 4 performs side channel attack resistance evaluation of the cryptographic device 1 using the side channel information subjected to signal processing.

  FIG. 12 is a flowchart of the side channel attack resistance evaluation means in Embodiment 11 of the present invention.

  The difference from the first embodiment is that after the side channel information is input in step A2, the characteristics of the input side channel information are calculated (step A9), and the parameters of the signal processing are determined according to the calculated characteristics. (Step A10).

  By performing signal processing using parameters according to the characteristics of the side channel information, unnecessary information can be removed more effectively than when signal processing is performed with fixed parameters. Further, since the signal processing unit 3 sets parameters according to the side channel information, it is possible to reduce the trouble of setting parameters in advance for each side channel information.

  Here, examples of features used for discrimination include statistics such as average value and median of side channel information, maximum value and minimum value, and variance, and the position and height of a peak on a power spectrum. The parameters are determined based on the above characteristics.

[Twelfth embodiment]
In the twelfth embodiment of the present invention to be described subsequently, the signal processor 3 calculates the power spectrum of the side channel information with respect to the side channel information measured by the encryption device 1 to be evaluated. Next, a signal processing parameter to be used is determined according to the calculated peak position of the power spectrum, and the signal processing is performed on the side channel information. Thereafter, the side channel attack resistance evaluation unit 4 performs side channel attack resistance evaluation of the cryptographic device 1 using the side channel information subjected to signal processing.

  FIG. 13 is a flowchart of the side channel attack resistance evaluation means in Embodiment 12 of the present invention.

  The difference from the eleventh embodiment is that, after the side channel information is inputted in step A2, the power spectrum of the inputted side channel information is calculated (step A9-1), and the power spectrum is calculated according to the peak position of the calculated power spectrum. Thus, the signal processing parameters are determined (step A10-1).

  For example, when determining the pass band of the low-pass filter as a parameter, a band less than or less than the peak position frequency can be adopted as the pass band.

  Similarly, when determining the pass band of the high-pass filter, a band higher than or higher than the peak position frequency can be used as the pass band.

  Further, when determining the pass band of the band pass filter, a band centered on the frequency at the peak position, a band less than the peak, and a band above the peak can be considered. In the case of the band centered on the peak, the lower limit of the band is 0, and the upper limit is the sampling frequency or Nyquist frequency of the measurement environment. In a band less than the peak, the lower limit of the band is 0 and the upper limit is the peak frequency. In the band above the peak, the lower limit of the band is the peak frequency, and the upper limit is the sampling frequency or Nyquist frequency of the measurement environment. Furthermore, it is conceivable to set the passband of the bandpass filter within the respective upper and lower limits.

  The peak of the power spectrum represents the characteristics of the component included in the side channel information, and by setting parameters with attention to the peak, it becomes easy to hold necessary information and remove unnecessary information. For example, when unnecessary information appears periodically in the side channel information, it appears as a peak with unnecessary information on the power spectrum. Therefore, unnecessary information can be attenuated by setting a pass band that avoids the peak.

  In addition, since the signal processing unit 3 sets parameters according to the side channel information, it is possible to reduce the trouble of setting parameters in advance for each side channel information.

[Thirteenth embodiment]
Next, an embodiment of the present invention will be described in which after the resistance evaluation is performed by the side channel attack resistance evaluation unit 4, the parameters used by the signal processing unit 3 are changed.

  FIG. 14 is a flowchart of the side channel attack resistance evaluation means in Embodiment 13 of the present invention.

  The difference from the first embodiment is that after performing the side channel attack resistance evaluation in step A4, if the evaluation is OK, it is checked whether there is room for change in the parameters applied in the signal processing (step A11). If the parameter can be changed, the parameter is changed (step A12), and the signal processing and the side channel attack resistance evaluation are performed again.

  By changing the parameters after evaluating the side channel attack resistance, depending on the content of the resistance evaluation (existence of resistance, correlation value and distance between side channel information and reference waveform, similarity, etc.) The parameter can be changed.

  For example, when the distance between the side channel information and the reference waveform is long, there is little possibility that the distance will be greatly reduced even if the parameter to be used is modified to be small. Therefore, the parameter correction amount is set so as to depend on the distance. Specifically, by setting the parameter to be corrected to a large value when the distance is long and to a small parameter when the distance is close, it is faster than correcting the parameter in small steps. An evaluation result can be obtained.

  In addition, since the evaluation is repeated until the parameter reaches the limit value or is determined to be NG in step A4, it can be erroneously determined to be resistant due to a parameter setting error even though it is actually not resistant. Sex can be reduced.

  Furthermore, since the signal processing unit 3 sequentially changes the parameters, the time and effort for setting the parameters in advance for each side channel information is reduced.

  In the twelfth embodiment of the present invention, an AES (Advanced Encryption Standard) is mounted on an evaluation board (corresponding to the encryption device 1) capable of performing encryption, and an oscilloscope (corresponding to the side channel information measurement device 2). Is used to measure an electromagnetic wave (corresponding to side channel information) leaked from the evaluation board during the encryption process, and evaluate the side channel attack resistance using the measured electromagnetic wave.

  In this embodiment, a step-pass filter is used for signal processing in step A3. Further, in step A4, a correlation value between the reference waveform and the side channel information is obtained, and when the correlation value is high, it is determined that the encryption device 1 does not have side channel attack resistance. Further, in step A10-1, the passband of the bandpass filter is determined as a parameter.

  First, a reference waveform used in step A4 is prepared. As the reference waveform, a waveform representing the characteristics of the first round of AES is used as a waveform representing the characteristics of AES. FIG. 15 shows a waveform representing the characteristics of AES acquired in advance. Here, the frame portion represents the first round of AES. Further, as a criterion for resistance evaluation, it is determined that there is no resistance when a plurality of peaks having a correlation value of 0.6 or more appear. The reason why the correlation value is used as a criterion for tolerance evaluation is that there is a correlation with the waveform representing the characteristics of the AES round because the timing at which the AES round processing is performed is determined from the side channel information. It is.

  Subsequently, AES is executed on the evaluation board, and the electromagnetic wave waveform of AES (FIG. 16) is measured (step A2).

  Next, the power spectrum of the AES electromagnetic wave waveform is obtained as a feature discrimination by the signal processing unit 3 for the measured AES electromagnetic wave waveform (step A9-1). FIG. 17 shows the power spectrum of the electromagnetic wave waveform of AES. In FIG. 17, the peak position is indicated by a circle.

  Subsequently, the passband of the bandpass filter, which is a parameter used in signal processing, is determined from the peak of the power spectrum obtained in step A9-1 (step A10-1). As a method of determining the pass band, a band centered on the peak, a band less than the peak, and a band above the peak can be considered.

  In the case of the band centered on the peak, the lower limit of the band is 0, and the upper limit is 1/2 of the sampling rate in the oscilloscope. In a band less than the peak, the lower limit of the band is 0 and the upper limit is the peak frequency. In the band above the peak, the lower limit of the band is the peak frequency, and the upper limit is ½ of the sampling rate.

  When the lower limit is 0, it is equivalent to a low-pass filter, and when the upper limit is 1/2 of the sampling rate, it is equivalent to a high-pass filter.

  In this embodiment, a band less than the peak is used and the passband width is 2 MHz. The upper limit of the pass band is a frequency that is 0.1 MHz away from the peak.

  Next, bandpass filter processing is performed on the electromagnetic wave waveform measured in step A2 using the parameters determined in step A12 (step A3). FIG. 18 shows the electromagnetic wave waveform after the bandpass filter processing.

  Subsequently, side channel attack resistance evaluation is performed from the electromagnetic wave waveform to which signal processing is applied in step A3 (step A4). As an evaluation, a correlation with a reference waveform prepared in advance is calculated. FIG. 19 shows the calculated correlation.

  When the calculated correlation is low, it can be determined that the confidential information is not leaked from the side channel information, and therefore it can be determined that the current evaluation target is resistant to side channel attacks (step A5).

  On the other hand, when the calculated correlation is high, it can be determined that the confidential information is leaked from the side channel information, and therefore it can be determined that the current evaluation target does not have side channel attack resistance (step A7). This is because a plurality of peaks having a correlation value of 0.6 or more appear this time.

  In the thirteenth embodiment of the present invention, A5 / 1 which is an encryption algorithm used in GSM (Global System for Mobile Communications) is mounted on an evaluation board (corresponding to the encryption device 1) that can execute encryption. This case will be described as an example.

  In this embodiment, an oscilloscope (corresponding to the side channel information measuring device 2) is used to measure power leaked from the evaluation board during encryption processing (corresponding to side channel information), and the measured power is used. Side channel attack resistance.

  First, a reference waveform for evaluation used by the side channel attack resistance evaluation unit 4 is created. In order to create a reference waveform, the A5 / 1 waveform is measured 100 times. Next, an average of the measured waveforms is obtained (FIG. 20), and median filter processing (window width 101) is applied to the average waveform (FIG. 21). In FIG. 21, the portion surrounded by a frame is used as a reference waveform. This part is an A5 / 1 LFSR (Linear Feedback Shift Register) operation process and an output process.

  In addition, as a criterion for resistance evaluation, it is determined that there is no resistance when a peak having a correlation value of 0.7 or more appears. The reason why the correlation value is used as a criterion for the tolerance evaluation is that there is a correlation with the waveform representing the A5 / 1 LFSR operation processing and output processing. From the side channel information, the timing when the above processing of A5 / 1 is performed is found. It is because it will do.

  Next, A5 / 1 is executed on the evaluation board, and the power waveform (FIG. 22) of A5 / 1 is measured (step A2).

  Next, the signal processing unit 3 performs signal processing on the measured power waveform of A5 / 1 (step A3). The signal processing unit 3 applies median filter processing with a window width of 3 as a parameter to remove unnecessary information. FIG. 23 shows a waveform of A5 / 1 when the median filter processing is applied with a window width of 3.

  Thereafter, the side channel attack resistance evaluation unit 4 calculates the correlation between the power waveform after applying the median filter and the reference waveform prepared in advance. The calculation results are shown in FIG. Subsequently, since the correlation of the calculation results does not indicate a high correlation, it is determined that the characteristics of the cryptographic processing on the evaluation board are not leaked (step A4, OK determination).

  Next, it is determined whether the window width, which is a median filter parameter, has reached a limit value (step A11). Here, the upper limit value of the window width is used as the limit value. The upper limit value of the window width is the number of data of the input power waveform, but the evaluator can arbitrarily set the upper limit value between 3 and the number of data of the power waveform. Here, when the window width reaches the upper limit value, since it is determined that the feature is not leaked, it is determined that the cryptographic apparatus 1 has side channel attack resistance (step A5).

On the other hand, if the window width has not reached the upper limit value, the window width is updated according to any of the following rules (step A12), and the process returns to step A3.
Rule (1) Increase the window width by 2.
Rule (2) The window width is increased according to the maximum value of the correlation calculated in the tolerance evaluation.

  Here, in the case of rule (1), the parameters are simply updated. On the other hand, in the case of rule (2), when the maximum value is small, even if the window width is increased by 2, it is unlikely that the correlation value will increase rapidly. Therefore, when the maximum value is small, the amount of update of the window width is increased, and when the maximum value approaches the reference value for resistance evaluation, the amount of update is decreased, so that the signal processing and the resistance can be improved as compared with the case of rule (1). It is possible to reduce the number of evaluations and determine resistance evaluation at high speed.

  In this embodiment, when the maximum value is less than 0.4, the update amount is 10, and when the maximum value is 0.4 or more and less than 0.65, the update amount is 6, or when the update amount is 0.65 or more and less than 0.7. Set the update amount to 2.

  FIG. 25 shows changes in the maximum value of the correlation when signal processing, tolerance evaluation, and parameter variation are repeated. Here, when the window width is updated according to the rule (1), the line indicates the place where the actual processing is performed when the window width is updated according to the rule (2).

  In this example, when the window width was 55, the maximum correlation value was 0.7 or more. Note that the number of processing steps A3 required to satisfy the standard is 27 in the case of rule (1), but 13 in the case of rule (2), and the number of processing is about half. It has become. Therefore, it can be said that the number of processes can be reduced by changing the parameter update amount according to the evaluation result.

  Next, FIG. 26 shows the A5 / 1 waveform when the median filter processing is applied when the window width is 55, and FIG. 27 shows the result when the correlation between FIG. 26 and the reference waveform is calculated.

  The correlation between the calculation results shows a high correlation, so that it can be determined that the characteristics of the cryptographic processing on the evaluation board are leaked (step A4, NG determination).

  For reference, FIG. 28 shows the result of calculating the correlation between the A5 / 1 power waveform (FIG. 22) acquired at step A2 and the reference waveform. Looking at the correlation of the calculation results, in the state where the median filter is not applied, the peak indicating the processing timing of A5 / 1 does not appear, and it is determined that there is side channel attack resistance. However, in practice, it is possible to detect that the processing timing is detected by applying a median filter and that the characteristics of the cryptographic processing on the evaluation board are leaked. Therefore, it can be seen that it is difficult to correctly evaluate the resistance of the encryption device from the waveform of FIG. 22 when the median filter is not applied.

  Note that the side channel attack resistance evaluation apparatus according to the embodiment of the present invention can be realized by hardware, software, or a combination thereof.

  This application is based on Japanese Patent Application No. 2007-319687 (filed on Dec. 11, 2007), and claims the priority of the Paris Convention based on Japanese Patent Application No. 2007-319687. The disclosure of Japanese Patent Application No. 2007-319687 is incorporated herein by reference to Japanese Patent Application No. 2007-319687.

  Although representative embodiments of the present invention have been described in detail, various changes, substitutions and alternatives may be made without departing from the spirit and scope of the invention as defined in the claims. It should be understood. Moreover, even if the claim is amended in the application procedure, the inventor intends that the equivalent scope of the claimed invention is maintained.

Claims (30)

In a side channel attack resistance evaluation device that evaluates resistance to side channel attacks performed using side channel information leaked from a cryptographic device,
Side channel information measuring means for measuring side channel information generated from the cryptographic device to be evaluated;
Signal processing means for performing signal processing for noise reduction on the side channel information obtained by the side channel information measuring means, and outputting the signal-processed side channel information;
Side channel attack resistance determination means for determining whether the evaluation target encryption device is resistant to side channel attacks based on the signal-processed side channel information output in the signal processing means and a reference waveform ;
A side channel attack resistance evaluation apparatus comprising: In the side channel attack tolerance evaluation apparatus according to claim 1,
The signal processing means is an average value filter based on an average value of side channel information, a median filter based on the median of side channel information, a frequency filter that determines a pass band based on the frequency of the peak position of the power spectrum of the side channel information, Alternatively, the side channel attack resistance evaluation apparatus characterized by performing signal processing using a multiband filter having an attenuation band in the vicinity of the frequency of the peak position of the power spectrum of the side channel information . In the side channel attack tolerance evaluation apparatus according to claim 2 ,
Wherein the frequency filter is a low pass filter to a frequency hereinafter the peak position of the power spectrum of the side-channel information the upper limit of the passband, high-pass filter having a higher frequency at the peak of the power spectrum of the lower limit of the passband side channel information, A bandpass filter whose passband is near the frequency of the peak position of the power spectrum of the side channel information, a notch filter whose attenuation band is near the frequency of the peak position of the power spectrum of the side channel information, or a passband which is the frequency of the side channel information. A side channel attack resistance evaluation apparatus, characterized in that it is a multiband filter having a frequency in the vicinity of the peak position of the power spectrum . In the side channel attack tolerance evaluation apparatus according to claim 1,
The side channel attack resistance evaluation apparatus, wherein the signal processing means performs signal processing using a filter based on a wavelet transform. In the side channel attack tolerance evaluation apparatus according to any one of claims 1 to 4 ,
The signal processing means includes means for calculating a feature quantity of the side channel information obtained by the side channel information measurement means and determining a parameter of signal processing to be performed based on the calculated feature quantity. Side channel attack resistance evaluation device. In the side channel attack tolerance evaluation apparatus according to claim 5 ,
A side channel attack resistance evaluation apparatus characterized in that a peak of a power spectrum of side channel information obtained by the side channel information measuring means is used as the feature amount. In the side channel attack tolerance evaluation apparatus according to any one of claims 1 to 6 ,
The side channel attack resistance evaluation apparatus, wherein the signal processing means includes means for changing a parameter of signal processing from the determination result of the side channel attack resistance determination means.   The side channel attack resistance evaluation device according to claim 7,
  When the intermediate determination result of the side channel attack determination means indicates that there is attack resistance, repeatedly changing the signal processing parameter so that a determination result that the attack resistance is weak is easily obtained, Even if the signal processing parameter reaches a predetermined limit value, if the intermediate determination result of the side channel attack determination means indicates that there is attack resistance, the final determination result is that there is attack resistance. Side channel attack resistance evaluation device.   The side channel attack resistance evaluation device according to claim 1,
  The signal processing means performs signal processing using a median filter based on the median of the side channel information;
  The signal processing means varies the window width of the median filter from the determination result of the side channel attack resistance determination means,
  When the intermediate determination result of the side channel attack determination means indicates that there is attack resistance, repeatedly changing the window width of the median filter so that the window width of the median filter is a predetermined limit value The side channel attack resistance evaluation device is characterized in that if the intermediate determination result of the side channel attack determination means indicates that there is attack resistance, the final determination result is that there is attack resistance.   The side channel attack resistance evaluation device according to claim 9,
  In the signal processing means calculated in the last resistance evaluation in the case where the intermediate determination result of the side channel attack determination means indicates that there is attack resistance and the window width of the median filter is changed to be wide. A side channel attack resistance evaluation apparatus characterized by increasing the window width according to the maximum value of the correlation between the output signal-processed side channel information and the reference waveform. In the side channel attack resistance evaluation method for evaluating resistance to side channel attacks performed using side channel information leaked from the cryptographic device,
A side channel information measurement step for measuring side channel information generated from the cryptographic device to be evaluated;
A signal processing step of performing signal processing for noise reduction on the side channel information obtained in the side channel information measurement step, and outputting the side channel information subjected to signal processing,
Side channel attack resistance determination step for determining whether or not the cryptographic device to be evaluated is resistant to side channel attacks based on the signal-processed side channel information output in the signal processing step and a reference waveform ;
A side channel attack resistance evaluation method comprising: In the side channel attack tolerance evaluation method according to claim 11 ,
The signal processing step includes an average value filter based on an average value of side channel information, a median filter based on the median of side channel information, a frequency filter that determines a pass band based on the frequency of the peak position of the power spectrum of the side channel information, Alternatively, the side channel attack resistance evaluation method, wherein signal processing is performed using a multiband filter having an attenuation band near the frequency of the peak position of the power spectrum of the side channel information . In the side channel attack tolerance evaluation method according to claim 12 ,
Wherein the frequency filter is a low pass filter to a frequency hereinafter the peak position of the power spectrum of the side-channel information the upper limit of the passband, high-pass filter having a higher frequency at the peak of the power spectrum of the lower limit of the passband side channel information, A bandpass filter whose passband is near the frequency of the peak position of the power spectrum of the side channel information, a notch filter whose attenuation band is near the frequency of the peak position of the power spectrum of the side channel information, or a passband which is the frequency of the side channel information. A side channel attack resistance evaluation method, which is a multiband filter having a frequency in the vicinity of a power spectrum peak position . In the side channel attack tolerance evaluation method according to claim 11 ,
The signal processing step performs signal processing using a filter based on a wavelet transform, wherein the side channel attack resistance evaluation method is characterized. In the side channel attack tolerance evaluation method according to any one of claims 11 to 14 ,
The signal processing step includes a step of calculating a feature amount of the side channel information obtained in the side channel information measurement step, and determining a parameter of signal processing to be performed based on the calculated feature amount. Side channel attack resistance evaluation method. The side channel attack resistance evaluation method according to claim 15 ,
A side channel attack resistance evaluation method, wherein a peak of a power spectrum of side channel information obtained in the side channel information measurement step is used as the feature amount. In the side channel attack tolerance evaluation method according to any one of claims 11 to 16 ,
The signal processing step includes a step of changing a parameter of signal processing from a determination result of the side channel attack resistance determination step.   The side channel attack resistance evaluation method according to claim 17,
  When the intermediate determination result of the side channel attack determination step indicates that there is attack resistance, repeatedly changing the signal processing parameter so that a determination result that the attack resistance is weak is easily obtained, Even if the signal processing parameter reaches a predetermined limit value, if the intermediate determination result of the side channel attack determination step indicates that there is attack resistance, the final determination result is that there is attack resistance. Side channel attack resistance evaluation method.   The side channel attack resistance evaluation method according to claim 11,
  The signal processing step performs signal processing using a median filter based on the median of the side channel information;
  The signal processing step varies the window width of the median filter from the determination result of the side channel attack resistance determination means,
  When the intermediate determination result of the side channel attack determination step indicates that there is attack resistance, repeatedly changing the window width of the median filter so that the window width of the median filter is a predetermined limit value The side channel attack resistance evaluation method is characterized in that, if the intermediate determination result of the side channel attack determination step indicates that there is attack resistance, the final determination result is that there is attack resistance.   The side channel attack resistance evaluation method according to claim 19,
  In the signal processing step, the intermediate determination result of the side channel attack determination step indicates that there is attack resistance, and when the median filter window width is changed to be widened, the signal processing step is calculated at the time of the previous resistance evaluation. A side channel attack resistance evaluation method, wherein the window width is increased in accordance with a maximum value of a correlation between the output signal-processed side channel information and a reference waveform. In a side channel attack resistance evaluation program that evaluates resistance to side channel attacks performed using side channel information leaked from the cryptographic device,
A side channel information measurement function for measuring side channel information generated from the cryptographic device to be evaluated;
A signal processing function for performing signal processing for noise reduction on the side channel information obtained in the side channel information measurement function, and outputting signal-processed side channel information;
Side channel attack resistance determination function for determining whether the evaluation target cryptographic device is resistant to side channel attacks based on the signal-processed side channel information output in the signal processing function and a reference waveform ;
Side channel attack resistance evaluation program characterized by causing a computer to realize In the side channel attack tolerance evaluation program according to claim 21,
The signal processing function is an average value filter based on an average value of side channel information, a median filter based on the median of side channel information, a frequency filter that determines a pass band based on the frequency of the peak position of the power spectrum of the side channel information, Alternatively, a side channel attack resistance evaluation program characterized in that signal processing is performed using a multiband filter having an attenuation band near the frequency of the peak position of the power spectrum of the side channel information. In the side channel attack resistance evaluation program according to claim 22,
The frequency filter is a low-pass filter whose upper limit of the pass band is equal to or lower than the frequency of the peak position of the power spectrum of the side channel information, and a high pass filter whose lower limit of the pass band is equal to or higher than the frequency of the peak position of the power spectrum of the side channel information. Bandpass filter whose band is near the frequency of the peak position of the power spectrum of the side channel information, notch filter whose attenuation band is near the frequency of the peak position of the power spectrum of the side channel information, or power of the side channel information whose passband is A side channel attack resistance evaluation program characterized by being a multiband filter having a frequency near the peak position of the spectrum. In the side channel attack tolerance evaluation program according to claim 21,
The side channel attack resistance evaluation program, wherein the signal processing function performs signal processing using a filter based on wavelet transform. In the side channel attack tolerance evaluation program according to any one of claims 21 to 24 ,
The signal processing function causes the computer to realize a function of calculating a feature amount of the side channel information obtained by the side channel information measurement function and determining a signal processing parameter to be performed based on the calculated feature amount. A characteristic side channel attack resistance evaluation program. In the side channel attack tolerance evaluation program according to claim 25,
A side channel attack resistance evaluation program characterized in that a peak of a power spectrum of side channel information obtained by the side channel information measurement function is used as the feature amount. In the side channel attack tolerance evaluation program according to any one of claims 21 to 26 ,
A side channel attack resistance evaluation program characterized in that the signal processing function causes a computer to realize a function of changing a parameter of signal processing from a determination result of the side channel attack resistance determination function.   The side channel attack resistance evaluation program according to claim 27,
  When the intermediate determination result of the side channel attack determination function indicates that there is attack resistance, repeatedly changing the signal processing parameter so that a determination result that the attack resistance is weak is easily obtained, Even if the signal processing parameter reaches a predetermined limit value, if the intermediate determination result of the side channel attack determination function indicates that there is attack resistance, the final determination result is that there is attack resistance. Side channel attack resistance evaluation program.   The side channel attack resistance evaluation program according to claim 21,
  The signal processing function performs signal processing using a median filter based on the median of the side channel information,
  The signal processing function varies the window width of the median filter from the determination result of the side channel attack resistance determination function,
  When the intermediate determination result of the side channel attack determination function indicates that there is an attack resistance, repeatedly changing the window width of the median filter so that the window width of the median filter is a predetermined limit value The side channel attack resistance evaluation program is characterized in that if the intermediate determination result of the side channel attack determination function indicates that there is attack resistance, the final determination result is that there is attack resistance.   A side channel attack resistance evaluation program according to claim 29,
  In the signal processing function, which is calculated in the last resistance evaluation, when the intermediate determination result of the side channel attack determination function indicates that there is attack resistance and the median filter window width is changed to be wide A side channel attack resistance evaluation program characterized by increasing the window width according to the maximum value of the correlation between the output signal processed side channel information and the reference waveform.

Images