Methods for maximizing the entropy derived from an analogous entropy source

Abstract

Method for generating a security identifier, wherein the method comprises the following steps: - Maximizing the entropy derived from an analog entropy source by a method for maximizing the entropy derived from an analog entropy source, wherein the method for maximizing the entropy derived from an analog entropy source comprises providing input data to the analog entropy source (2), generating return values ​​by the analog entropy source based on the input data (3), and grouping the return values, wherein the grouping of the return values ​​comprises applying offsets to the return values, and - Generating the security identifier based on the generated maximized entropy (8), wherein the analog entropy source is a Physical Unclonable Function or a biometric source, and wherein applying offsets divides return values ​​into the same group, even though the distance between the corresponding return values ​​is less than a corresponding limit for the distance between two return values, by introducing offsets so that it appears as if the distance between the corresponding return values ​​corresponds to the corresponding limit for the distance between two return values.

Description

[0001] The present invention relates to a method for maximizing the entropy derived from an analog entropy source, which method is designed such that a comparatively high entropy can be achieved with comparatively few computer resources.

[0002] For example, in the Internet of Things (IoT), where many objects, such as sensors, actuators, devices, and services, are communicatively linked, security solutions are used to ensure the security of communication channels and to verify the identity and authorizations of objects. Many of these security solutions are based on cryptographic methods that rely on security identifiers, such as cryptographic keys, stored locally on objects. However, the fact that the security identifiers are stored locally on objects represents an increased security risk, especially since objects within an Internet of Things environment are often not physically secured.

[0003] One potential solution to this problem is entropy sources, such as physical unclonable functions. Entropy is a measure that indicates the average information content of messages generated for a message source. Physical unclonable functions are also hardware structures in a semiconductor that serve to enable unique identification of the semiconductor or to secure keys for cryptographic procedures. They are based on the fact that even the smallest fluctuations in the production process cause certain components to exhibit individual behavior, even though a uniform production process should produce absolutely identical parts. This leads to physical unclonable functions processing input data in such a way that different physical unclonable functions each output different and unique return values ​​based on the same input data.

[0004] To increase the generated entropy, the return values ​​of analog entropy sources can be divided into groups. However, this has the disadvantage that differences or errors in the generated return values, such as swapping of return values, can occur, particularly due to noise and / or different environmental conditions. Consequently, when grouping the return values, care must be taken to ensure that the distance between any two return values ​​assigned to the same group corresponds to at least a correspondingly specified limit for the distance between two return values. This prevents the corresponding group from being assigned a swapped return value instead of the expected one, and thus avoids bit errors during the subsequent generation of a security identifier.This in turn results in restrictions on the grouping of the return values ​​and thus also on the achievable entropy, whereby the length of the corresponding cryptographic key, which influences the forgery security of the corresponding key, and thus also on the achievable security, is based on the entropy ultimately derived from the corresponding entropy source.

[0005] However, known methods for increasing the entropy derived from an analog entropy source are also very resource-intensive, so that they can hardly be carried out on ordinary electronic devices, such as a personal computer or a laptop, which have limited resources.

[0006] Overall, there is a need for methods that can maximize the entropy derived from an analog entropy source.

[0007] From the document US 10,965,475 B2 a method for generating a security key for an integrated circuit device is known, which comprises generating a plurality of key bits with a generator for physical unclonable functions (PUF), wherein unstable bits of the plurality of key bits are identified, and a security key is generated based on the plurality of key bits, wherein the security key excludes the identified unstable bits.

[0008] From the paper Abdolinezhad, Saeed et al: “A Novel Key Generation Method for Group-Based Physically Unclonable Function Designs” a method for optimizing the entropy derived from an analog entropy source is known.

[0009] From the paper Yin, Chi-En Daniel et al.: “Maximizing RO PUF's secret extraction” a method for grouping return values ​​based on the longest increasing subsequence-based grouping algorithm is known, which is used to optimize hardware utilization.

[0010] From the paper Yin, Chi-En et al: “Design and Implementation of a Group-based RO PUF” a method for grouping return values ​​based on frequency measurements is known.

[0011] The application of spatial autocorellation analysis for physical unclonable functions is known from the publication Wilde, Florian et al: “Spatial Corellation Analysis on Physical Unclonable Functions”.

[0012] From the document US 10,230,369 B2 a method for generating a bit string based on a physical unclonable function is known, which is based on measured delays.

[0013] The object of embodiments of the invention is to provide an improved method for generating a security identifier.

[0014] This problem is solved by the subject matter of the subordinate claims. Further advantageous developments are the subject matter of the dependent claims.

[0015] According to one embodiment, this object is achieved by a method for maximizing the entropy derived from an analog entropy source, the method comprising providing input data to the entropy source, generating return values ​​by the entropy source based on the input data, and grouping the return values, wherein the grouping of the return values ​​comprises applying offsets to return values.

[0016] An offset is understood as a shift or deviation from an original position or a standard position. Applying offsets to return values ​​thus means that at least some of the return values ​​are offset or shifted from their original position, for example, their original position in a coordinate system or on an axis describing the return values, on which the return values ​​are sequentially labeled based on their value.

[0017] Shifting return values ​​accordingly has the advantage that the corresponding return values ​​can be shifted in such a way that the distance between two return values ​​after shifting theoretically corresponds to at least the corresponding limit for the distance between two return values, even if the distance between the same return values ​​was originally smaller than the corresponding limit. This, in turn, results in the generation of larger groups of return values ​​or groups with more elements, which also maximizes the generated entropy.

[0018] Since only shifts of the original positions of the corresponding return values ​​need to be considered, the method can also be carried out with comparatively few or limited computer resources, such as memory or processor capacities, and thus also on conventional electronic devices.

[0019] Overall, a method is provided with which the entropy derived from an analog entropy source can be maximized with comparatively few computer resources.

[0020] In one embodiment, the step of grouping the return values ​​comprises determining anchor points such that the distance between two immediately consecutive anchor points corresponds in each case to a limit value for the distance between two return values, and dividing the return values ​​into groups by successively assigning a return value to each of the anchor points based on at least one assignment rule, wherein the corresponding return value is spaced from the corresponding anchor point by less than the limit value for the distance between two return values.

[0021] An anchor point is understood to be a point that is located at a predetermined or fixed position, for example a fixed position in a coordinate system or on an axis describing the return values. The fact that the distance between two immediately consecutive anchor points corresponds to a limit value for the distance between two return values ​​means that the anchor points are distributed so equidistantly within the coordinate system or along the axis describing the return values ​​that a distance between two immediately consecutive anchor points is always equal to the limit value for the distance between two return values. It should be noted that the first anchor point can be freely chosen and, theoretically, can also coincide with the first return value.

[0022] An assignment rule is also understood to be a rule based on which a return value is moved or mapped to the corresponding anchor point. For example, the assignment rule can specify that the return value which is the shortest distance from an anchor point on an axis describing the return values ​​is mapped or moved to the corresponding anchor point. Furthermore, the assignment rule can also specify, for example, that the return value with the shortest distance to an anchor point among the return values ​​following the corresponding anchor point on an axis describing the anchor points is mapped to the corresponding anchor point, or that the return value with the shortest distance to an anchor point among the return values ​​lying before the corresponding anchor point on an axis describing the anchor points is mapped to the corresponding anchor point.

[0023] Overall, the step of grouping the return values ​​can thus be carried out in a simple manner with comparatively few computer resources.

[0024] In particular, a group can be formed as soon as exactly one return value has been assigned to each anchor point, whereby the corresponding return values ​​assigned to the anchor points form the elements of the corresponding group.

[0025] In addition, for each offset, the method may further comprise a step of storing the offset together with an identifier of the corresponding return value.

[0026] An identifier is an artificially assigned characteristic for the unique identification of an object.

[0027] This ensures that the offsets can be accessed directly and easily and taken into account, for example, if the elements of one or more of the generated groups of return values ​​are to be further processed, for example to generate a cryptographic key, without the corresponding offsets having to be derived first.

[0028] The analog entropy source is also a Physical Unclonable Function or a biometric source.

[0029] A biometric source is understood to be a unit that is designed to capture biometric characteristics that exhibit entropy, for example with the help of a fingerprint sensor or scanner.

[0030] The method can thus be based on known and common analog entropy sources without the need for complex and costly changes and / or adaptations.

[0031] According to the invention, a method for generating a security identifier is also provided, the method comprising maximizing the entropy derived from an analog entropy source by a method described above for maximizing the entropy derived from an analog entropy source, and generating the security identifier based on the generated maximized entropy.

[0032] A security identifier is understood to be a unique, unchangeable identifier of a user or object, a user or object group, or another security principle, for example a cryptographic key.

[0033] Thus, a method for generating a security identifier is provided, which is based on a method that maximizes the entropy derived from an analog entropy source with comparatively few computing resources. The fact that the security identifier, for example, a cryptographic key, can be determined based on increased entropy also leads to a corresponding increase in the length of the security identifier, for example, the cryptographic key, and thus also the achievable security.

[0034] A further embodiment of the invention also provides a method for identifying an object, wherein the object is identified based on a security identifier, and wherein the security identifier was generated by a method for generating a security identifier as described above.

[0035] Thus, a method for identifying an object is provided, which is based on a method that maximizes the entropy derived from an analog entropy source with comparatively few computing resources. In particular, the method is based on a security identifier, for example, a cryptographic key, which was determined based on increased entropy, so that the length of the corresponding security identifier, for example, the cryptographic key, and thus the achievable security, could also be increased accordingly.

[0036] A further embodiment also provides a system for maximizing the entropy derived from an analog entropy source, the system comprising a providing unit configured to provide input data for an analog entropy source, an analog entropy source configured to generate return values ​​based on the input data, and a grouping unit configured to group the return values, wherein the grouping of the return values ​​comprises applying offsets to return values.

[0037] Thus, a system is provided with which the entropy derived from an analog entropy source can be maximized with comparatively few computing resources. The fact that the system is designed to shift the corresponding return values ​​accordingly has the advantage that the return values ​​can be shifted in such a way that the distance between two return values ​​after shifting theoretically corresponds at least to the corresponding limit for the distance between two return values, even though the distance between the same return values ​​was originally smaller than the corresponding limit. This, in turn, results in the generation of larger groups of return values ​​or groups containing more elements, which also maximizes the generated entropy.Since only shifts of the original positions of the corresponding return values ​​need to be considered, the system can also have comparatively few or relatively low computer resources, such as memory or processor capacities.

[0038] In one embodiment, the grouping unit comprises a determination unit configured to determine anchor points such that the distance between two immediately consecutive anchor points corresponds to a limit value for the distance between two return values, and a classification unit configured to classify the return values ​​into groups by successively assigning a return value to each of the anchor points based on at least one assignment rule, wherein the corresponding return value is spaced from the anchor point by less than the limit value for the distance between two return values. Overall, the grouping unit can thus be configured to carry out the grouping of the return values ​​in a simple manner using comparatively few computing resources.

[0039] In this case, a group can be formed in particular as soon as exactly one return value has been assigned to each anchor point, whereby the corresponding return values ​​assigned to the anchor points form the elements of the corresponding group.

[0040] In addition, the system can further comprise a memory configured to store the offset for each offset along with an identifier of the corresponding return value. This ensures that the offsets can be accessed directly and easily and taken into account, for example, if the elements of one or more of the generated groups of return values ​​are subsequently to be further processed, for example, to generate a cryptographic key, without first having to derive the corresponding offsets.

[0041] The analog entropy source is, in turn, a Physical Unclonable Function or a biometric source. The system can thus be based on known and common analog entropy sources without the need for complex and costly modifications and / or adaptations.

[0042] According to the invention, a system for generating a security identifier is also provided, the system comprising a system as described above for maximizing the entropy derived from an analog entropy source and a generation unit configured to generate a security identifier based on a maximized entropy generated by the system for maximizing the entropy derived from an analog entropy source.

[0043] Thus, a system for generating a security identifier is provided, which is based on a system that maximizes the entropy derived from an analog entropy source. The fact that the security identifier, for example, a cryptographic key, can be determined based on increased entropy also leads to a corresponding increase in the length of the security identifier, for example, the cryptographic key, and thus also in the achievable security.

[0044] A further embodiment of the invention also provides a system for identifying an object, wherein the object is identified based on a security identifier, and wherein the security identifier was generated by a system for generating a security identifier as described above.

[0045] Thus, a system for identifying an object is provided, which is based on a system that maximizes the entropy derived from an analog entropy source. In particular, the system is based on a security identifier, for example, a cryptographic key, which was determined based on increased entropy, so that the length of the corresponding security identifier, for example, the cryptographic key, and thus the achievable security, could be increased accordingly.

[0046] In summary, the present invention provides a method for maximizing the entropy derived from an analog entropy source, which is designed such that a comparatively high entropy can be achieved with comparatively few computer resources.

[0047] The invention will now be explained in more detail with reference to the attached figures. Fig. 1 shows a flowchart of a method for identifying an object according to embodiments of the invention; Fig. 2a-d illustrate the step of grouping the return values ​​according to a first embodiment; Fig. 3 shows a schematic block diagram of a system for maximizing the entropy derived from an analog entropy source according to embodiments of the invention.

[0048] Fig. 1 shows a flowchart of a method for identifying an object 1 according to embodiments of the invention.

[0049] With an analog entropy source, it is important to first digitize the return values ​​generated by the analog entropy source in order to be able to process them further and, for example, generate a cryptographic key.

[0050] For example, if an entropy source generates n return values ​​based on n input values ​​or n input data, with a security identifier to be generated based on the return values, it seems most obvious that the corresponding digitization step involves a pairwise comparison of the n return values ​​and a binary bit is generated based on each of the corresponding comparisons. A corresponding security identifier thus has n / 2 binary bits.

[0051] Furthermore, with n return values, it is also conceivable to compare the n output values ​​simultaneously, for example, to compare the amplitudes of the individual return values, which leads to n! arrangement possibilities, which in turn can generate a significantly longer security identifier. However, the individual generated bits are independent of each other, so that corresponding methods for generating a security identifier must take into account both the length of the security identifier and a certain randomness of the individual bits.

[0052] More complex comparisons of the return values ​​lead to longer security identifiers, whereby it has proven advantageous to sort or group the return values ​​generated by an analog entropy source in groups in order to obtain the highest possible entropy and to be able to generate correspondingly long security identifiers.

[0053] However, this has proven to be disadvantageous in that differences or errors in the generated return values, such as a swapping of return values, can occur, particularly due to noise and / or different environmental conditions. Consequently, when grouping the return values, care must be taken to ensure that the distance between any two return values ​​assigned to the same group corresponds to at least a correspondingly specified limit value for the distance between two return values. This prevents the return value from being assigned to the corresponding group instead of the expected one, and thus avoids bit errors during the subsequent generation of a security identifier.This in turn results in restrictions on the grouping of the return values ​​and thus also on the achievable entropy, whereby the length of the corresponding cryptographic key, which influences the forgery security of the corresponding key, and thus also on the achievable security, is based on the entropy ultimately derived from the corresponding entropy source.

[0054] However, known methods for increasing the entropy derived from an analog entropy source are also very resource-intensive, so that they can hardly be carried out on ordinary electronic devices, such as a personal computer or a laptop, which have limited resources.

[0055] Overall, there is a need for methods that can maximize the entropy derived from an analog entropy source.

[0056] Fig. 1 shows a method 1 which comprises a step 2 of providing input data for the entropy source, a step 3 of generating return values ​​by the entropy source based on the input data and a step 4 of grouping the return values, wherein the grouping of the return values ​​comprises applying offsets to return values.

[0057] Shifting return values ​​accordingly has the advantage that the corresponding return values ​​can be shifted in such a way that the distance between two return values ​​after shifting theoretically corresponds to at least the corresponding limit for the distance between two return values, even if the distance between the same return values ​​was originally smaller than the corresponding limit. This, in turn, results in the generation of larger groups of return values ​​or groups with more elements, which also maximizes the generated entropy.

[0058] Since only shifts of the original positions of the corresponding return values ​​need to be considered, the method can also be carried out with comparatively few or limited computer resources, such as memory or processor capacities, and thus also on conventional electronic devices.

[0059] In particular, the method may comprise analyzing the generated return values, wherein analyzing the return values ​​comprises deciding which return values ​​can be combined into a group, and wherein deciding comprises applying offsets to the return values.

[0060] Overall, a method 1 is presented with which the entropy derived from an analog entropy source can be maximized with comparatively few computer resources.

[0061] According to the invention, a method 1 is thus provided in which return values ​​can be assigned to the same group even though the distance between these return values ​​is smaller than the corresponding limit value for the distance between two return values, by introducing offsets or auxiliary data so that it appears as if a distance between these return values ​​corresponds to the limit value for the distance between two return values.

[0062] For further processing of the groups of return values, the individual offsets or auxiliary data, i.e. the corresponding errors introduced, can then be calculated out again.

[0063] According to the embodiments of the Fig. 1, step 4 of grouping the return values ​​comprises a step 5 of defining anchor points such that the distance between two immediately consecutive anchor points corresponds to a limit value for the distance between two return values, and a step 6 of dividing the return values ​​into groups by successively assigning a return value to each of the anchor points based on at least one assignment rule, wherein the corresponding return value is spaced from the corresponding anchor point by less than the limit value for the distance between two return values. A group can be formed in particular as soon as exactly one return value has been assigned to each anchor point, wherein the corresponding return values ​​assigned to the anchor points form the elements of the corresponding group.

[0064] Step 6 of dividing the return values ​​into groups is thus based on calculating distances between anchor points and return values, with the elements of each group being determined based on the calculated distances and the applied or selected at least one assignment rule. It can also be specified that in step 6, only the distance to the nearest anchor point is calculated for each return value and is included in the determination of the elements of the individual groups.

[0065] In addition, the return values ​​can also be regenerated and / or remeasured after the step of setting anchor points has been performed and before step 6 of dividing the return values ​​into groups is performed.

[0066] How Fig. 1, the method 1 further comprises a step 7 of storing, for each offset, the offset together with an identifier of the corresponding return value.

[0067] The data can be stored in a non-volatile memory, for example in the form of a look-up table, without having to take major security requirements into account.

[0068] According to the embodiments of the Fig. 1, the analog entropy source is also a Physical Unclonable Function, which can, for example, be designed to react individually and uniquely to an incidence of light.

[0069] The generated maximized groups or the corresponding maximized entropy can then be used, for example, to generate security identifiers, such as cryptographic keys, which can be used, for example, to verify authorizations or to encrypt messages within an Internet of Things.

[0070] As can be seen, the method 1 further comprises a step 8 of generating a security identifier based on the generated maximized entropy.

[0071] The generation of the security identifier in step 8 can be based on known methods for generating a security identifier.

[0072] In addition, in step 8, the elements of each group can first be ordered based on the corresponding values, whereby a security identifier can then be generated, for example, based on a corresponding permutation of the elements of the group.

[0073] Typically, a separate security identifier is generated based on each generated group, although it is also possible to subsequently link the individual security identifiers to create a comparatively long security identifier.

[0074] Furthermore, during an introduction or commissioning phase, the individual groups can be created and the corresponding offsets can be calculated and saved, whereby only the return values ​​are measured for the subsequent generation of a safety identifier and otherwise the saved values ​​can be used with regard to the offsets.

[0075] According to the embodiments of the Fig. 1, the method 1 further comprises a step 9 of identifying an object based on the generated security identifier.

[0076] The object can, for example, be an object within an Internet of Things environment or a message sent within an Internet of Things.

[0077] The Fig. 2a - d illustrate step 10 of grouping the return values ​​according to a first embodiment.

[0078] In particular, the Fig. 2a - d is an axis describing the return values, on which the return values ​​11 are entered in sequence based on their values.

[0079] Fig. 2a shows additional anchor points 12, 13 entered on the axis, wherein the distance between two immediately consecutive anchor points 12, 13 corresponds to a limit value for the distance between two return values ​​11, and wherein a first anchor point 13 corresponds to a first return value.

[0080] The Fig. 2b - d further show the division of the return values ​​11 into groups by successively assigning a return value 11 to each anchor point 12, 13 based on an assignment rule, wherein the corresponding return value 11 is spaced from the corresponding anchor point 12, 13 by less than the limit value for the distance between two return values.

[0081] According to the first embodiment, the assignment rule states that the return value is assigned to each anchor point, i.e. theoretically shifted to the anchor point which has the smallest distance to the corresponding anchor point.

[0082] Fig. Figure 2b shows the formation of a first group of return values ​​by selecting the return values ​​14 which, among all return values ​​11, have the shortest distance to one of the anchor points 12, 13 as elements of the first group.

[0083] Fig. 2c further shows the formation of a second group of return values ​​by selecting from the remaining return values ​​the return values ​​15 which have the shortest distance to one of the anchor points 12, 13 among all remaining return values ​​as elements of the second group.

[0084] Fig. 2d further shows the formation of a third group of return values ​​from the remaining return values ​​based on the same assignment rule, however, no group can be formed from the last remaining return value 16 since, according to the first embodiment, a group of return values ​​should always have at least two elements or return elements.

[0085] In other embodiments, the assignment rule may, for example, specify that when dividing the group, the return value is assigned to an anchor point that has the shortest distance to the corresponding anchor point and lies within a search radius around the anchor point. The search radius may, for example, correspond to the threshold value for the distance between two return values.

[0086] Fig. 3 shows a schematic block diagram of a system 20 for maximizing the entropy derived from an analog entropy source according to embodiments of the invention.

[0087] How Fig. 3, the system comprises a providing unit 21 configured to provide input data for an analog entropy source, an analog entropy source 22 configured to generate return values ​​based on the input data, and a grouping unit 23 configured to group the return values, wherein the grouping of the return values ​​comprises applying offsets to return values.

[0088] The supply unit can be, for example, a light source or another signal source suitable for providing signals that can be processed by an analog entropy source. The grouping unit can also be implemented, for example, based on code stored in a memory and executable by a processor.

[0089] As can be seen, the grouping unit 23 according to the embodiments of the Fig. 3 further comprises a determination unit 24 which is designed to determine anchor points in such a way that the distance between two immediately consecutive anchor points corresponds in each case to a limit value for the distance between two return values, and a classification unit 25 which is designed to classify the return values ​​into groups by successively assigning a return value to each of the anchor points based on at least one assignment rule, wherein the corresponding return value is spaced from the corresponding anchor point by less than the limit value for the distance between two return values.

[0090] Further shows Fig. 3 a memory 26 configured to store the offset for each offset along with an identifier of the corresponding return value. This is again a non-volatile memory, and the data can be stored, for example, in the form of a look-up table.

[0091] According to the embodiments of the Fig. 3, the analog entropy source is again a Physical Unclonable Function.

Claims

[1] A method for generating a security identifier, the method comprising the following steps: - Maximizing the entropy derived from an analog entropy source by a method for maximizing the entropy derived from an analog entropy source, wherein the method for maximizing the entropy derived from an analog entropy source comprises providing input data to the analog entropy source (2), generating return values ​​by the analog entropy source based on the input data (3), and grouping the return values, wherein the grouping of the return values ​​comprises applying offsets to return values, and - generating the security identifier based on the generated maximized entropy (8), wherein the analog entropy source is a physical unclonable function or a biometric source, and wherein applying offsets comprises classifying return values ​​into the same group even though the distance between the corresponding return values ​​is less than a corresponding threshold for the distance between two return values, by introducing offsets such that the distance between the corresponding return values ​​appears to correspond to the corresponding threshold for the distance between two return values. [2] The method of claim 1, wherein the step of grouping the return values ​​(4) further comprises the following steps: - setting anchor points such that the distance between two immediately consecutive anchor points corresponds to a limit value for the distance between two return values ​​(5); and - dividing the return values ​​into groups by successively assigning a return value to each of the anchor points based on at least one assignment rule, wherein the corresponding return value is spaced from the corresponding anchor point by less than the limit value for the distance between two return values ​​(6). [3] A method according to claim 1 or 2, wherein the method further comprises the following step: - For each offset, store the offset together with an identifier of the corresponding return value (7). [4] A method for identifying an object, wherein the object is identified based on a security identifier, and wherein the security identifier was generated by a method for generating a security identifier according to any one of claims 1 to 3. [5] A system for generating a security identifier, the system comprising a system for maximizing the entropy derived from an analog entropy source and a generation unit configured to generate a security identifier based on a maximized entropy generated by the system for maximizing the entropy derived from an analog entropy source, the system for maximizing the entropy derived from an analog entropy source (20) comprising a providing unit (21) configured to provide input data for an analog entropy source, an analog entropy source (22) configured to generate return values ​​based on the input data, and a grouping unit (23) configured to group the return values, the grouping of the return values ​​comprising applying offsets to return values,wherein the analog entropy source is a Physical Unclonable Function or a biometric source, and wherein applying offsets comprises classifying return values ​​into the same group even though the distance between the corresponding return values ​​is less than a corresponding limit on the distance between two return values, by introducing offsets such that the distance between the corresponding return values ​​appears to correspond to the corresponding limit on the distance between two return values, [6] System according to claim 5, wherein the grouping unit (23) comprises a determination unit (24) which is designed to determine anchor points such that the distance between two immediately consecutive anchor points corresponds in each case to a limit value for the distance between two return values, and a classification unit (25) which is designed to classify the return values ​​into groups by successively assigning a return value to each of the anchor points based on at least one assignment rule, wherein the corresponding return value is spaced from the corresponding anchor point by less than the limit value for the distance between two return values. [7] System according to claim 5 or 6, wherein the system further comprises a memory (26) which is designed to store, for each offset, the offset together with an identifier of the corresponding return value. [8] A system for identifying an object, wherein the object is identified based on a security identifier, and wherein the security identifier was generated by a system for generating a security identifier according to any one of claims 5 to 7.

Images