Method for detecting an anomaly in an electronic system in operation
Patent Information
- Application Number
- EP2022840223
- Authority / Receiving Office
- EP · EP
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2021-12-31
- Filing Date
- 2022-12-30
- Publication Date
- 2024-11-06
Smart Images

Figure 1.1
Abstract
Description
[0001] DESCRIPTION
[0002] TITLE: Method for detecting an anomaly in an electronic system in operation
[0003] The present invention relates to a method for detecting an anomaly in an electronic system in operation, in particular a computer attack against the electronic system.
[0004] The invention applies in particular to the field of critical on-board computers such as those present in avionics, space, railways, automobiles, maritime, medical or IoT ("Industrial Internet of Things" in English or "Industrial Internet of Things" in French), traditionally confronted with operational safety issues, the role of which is to protect these systems and their users from failures and breakdowns.
[0005] Numerous certification standards related to these industries, such as the DO178C / DO254 standard for avionics, the IEC61508 standard for the industrial sector, the EN5012x standard for the railway sector, and the ECSSx standard for the space sector, provide the necessary safety guarantees on the reliability and availability of these systems. However, systems are becoming increasingly connected, which introduces new issues related to cybersecurity and resistance to malicious attacks.
[0006] For example, in the avionics field, it is planned that air traffic control, which has a better overview, could pilot aircraft from the ground during the approach phases to airports, introducing new issues of authentication, confidentiality and integrity.
[0007] The IoT takes the principles of the Internet of Things (or "IoT") and applies them to the industrial sector, sharing data sensors from different systems and maximizing machine-to-machine communications without human intervention. The IoT promotes constant communication between different systems to increase operational efficiency, but it faces the need to secure these communications and ensure the security of each of the systems that make it up, so that it cannot be used as a backdoor to enter the system maliciously. The field of wearable computing and implanted medical systems, in addition to classic security issues, also faces issues of confidentiality and data integrity.
[0008] Furthermore, secure systems are traditionally updated as soon as a security vulnerability is discovered and fixed. However, systems with a high level of operational reliability are rarely updated, as this generally involves recertifying the entire system, which can have very high financial costs. Effective security of critical embedded systems must therefore not only rely on ad-hoc defenses for each type of attack, each requiring an update, but must also enable the proactive detection of new malicious attacks.
[0009] In cybersecurity, we know about intrusion detection systems (IDS). These systems are generally classified into two categories: NIDS (Network Intrusion Detection System) and HIDS (Host Intrusion Detection System). NIDS focuses on analyzing network communications. It involves capturing traffic at a particular point in a network and analyzing the packets by comparing them to a database of attack patterns. If a set of packets matches a known attack pattern, an alert is raised. HIDS focuses on a computer, at the level of its operating system. It involves detecting abnormal behavior by monitoring running processes, memory allocations, logged-in users, etc.An alert is generated when one of the monitored quantities deviates from a predefined standard.
[0010] All of these systems focus on security issues at either the software or communications level. However, more recent attacks have begun to exploit hardware vulnerabilities at the processor level, which are much more complex and costly to patch.
[0011] For example, the Spectre attack exploits a hardware vulnerability in some implementations of branch prediction on microprocessors with speculative execution. While at the functional level, a misprediction is rolled back as soon as it is detected, speculative execution can, however, as a side effect, allow arbitrary access to RAM locations, with data used by the code that was executed "just in case" remaining stored in the cache.
[0012] The Meltdown attack also exploits a vulnerability related to speculation used for out-of-order execution of instructions in the processor, including cache memories and the TLB (Translation Lookaside Buffer). Unlike the Spectre attack, the Meltdown attack uses mechanisms more specific to the processor architecture, and allows access to data without the corresponding access rights, but also privilege escalation.
[0013] The Rowhammer attack exploits an unintended side effect of dynamic random access memory (DRAM) that causes electrical charge to leak into adjacent memory cells, allowing the contents stored in these neighboring cells to be modified without having access rights.
[0014] Thus, known detection methods have become insufficient in the face of this new generation of computer attacks.
[0015] There is therefore a need for a method capable of better detecting and protecting against anomalies affecting the electronic system, in particular computer attacks.
[0016] For this purpose, a method is proposed for detecting an anomaly in an electronic system in operation, in particular a computer attack against the electronic system, the electronic system comprising one or more processors each composed of a plurality of hardware blocks, the processor being adapted to execute at least one application by interactions of the application with said hardware blocks, the detection method comprising at least the following steps:
[0017] - measurement, during operation of the electronic system, of at least one parameter representative of the interactions of the application with one of the hardware blocks to obtain measurements of said representative parameter,
[0018] - for each measurement, comparison of said measurement with an associated reference data set to detect any inconsistencies between the measurement and the reference data set, said reference data set being representative of the operation of the electronic system without anomalies, and - issuing an alert according to an alert criterion relating to the inconsistency(ies) detected.
[0019] According to other embodiments, the method comprises one or more of the following characteristics, taken in isolation or in all technically possible combinations:
[0020] - the detection method further comprises at least one step chosen from the group consisting of: analysis of the nature of the hardware block associated with each inconsistency, analysis of the temporal frequency of the inconsistencies, and analysis of the deviation of each measurement associated with the inconsistencies from the reference data set; the detection method comprising a detection step, depending on the or each analysis of an anomaly, in particular a computer attack or a hardware failure of one of the hardware blocks.when at least one alert associated with one of the hardware blocks is issued, a new iteration of the steps of the detection method from the measurement step is implemented, the set of hardware blocks for which a representative parameter measurement is carried out being, at each iteration, included in the set of hardware blocks of the previous iteration, the set of hardware blocks being in particular included in the set of hardware blocks associated with the alert(s). each alert criterion depends on at least one condition on the representative parameters, the detection method comprising, when an alert associated with one of the hardware blocks is issued, a new iteration of the steps of the detection method from the measurement step, the alert criterion comprising, at each iteration, an increasing number of conditions.each representative parameter associated with one of the hardware blocks is chosen from the group consisting of: the type of data processed by the hardware block; the type of instructions executed by the hardware block; the number of accesses to the hardware block; the number of calculations executed and / or the calculation time of the hardware block; the number of malfunctions of the hardware block, and the number of interactions on an interconnection network internal to the processor involving the hardware block, and the number of interactions with the outside of the processor by the hardware block.
[0021] - each hardware block is chosen from the group consisting of: one or more calculation units, one or more branch prediction units, one or more internal memory registers of the processor, one or more cache memories, one or more random access memories, one or more processing chains, one or more memory protection or address translation units, one or more interconnection buses or networks, one or more input-output units each parameter representative of the operation of one of the hardware blocks being chosen from the group consisting of: the number of data of the same type processed by the calculation unit(s), the interconnection network(s) or the memory(s), the number of instructions of the same type executed by the calculation unit(s), the energy consumption of the calculation unit(s), the temperature of the calculation unit(s),the number of successes and / or errors in prediction of the branch prediction unit, the number of accesses to memories, the number of successes and / or errors in requests to the cache memory, the number of successes and / or errors in requests to the memory protection unit, the execution time of an instruction in the processing chain, the number of exchanges of information via the interconnection network(s) internal to the processor, the number of exchanges of information via the interconnection network(s) obtained by a given source, the number of exchanges of information via the interconnection network(s) sent to a given destination, and the number of exchanges of information via the input-output unit.,
[0022] - the electronic system comprises a plurality of hardware counting blocks each configured to measure one of the representative parameters, the measuring step comprising the measurement of a plurality of representative parameters, in particular between four and six, for example the representative parameters measured by a portion of the hardware counting blocks.
[0023] - the detection method further comprises a step of: comparing the or each inconsistency associated with the alert issued with a set of predetermined known anomalies, and categorizing the alert when the or each inconsistency corresponds to one of the known anomalies of said set of anomalies.
[0024] - the electronic system is a computer embedded in a transport platform. the detection method is implemented by a system chosen from the group consisting of: a dedicated programmable logic circuit forming one of said hardware blocks of the processor; software of an operating system of the electronic system configured to interact with a memory protected by a memory protection unit, and an application capable of being implemented by the electronic system by adding a driver controlling the hardware blocks.
[0025] Also provided is a computer program product comprising a readable information medium, on which is stored a computer program comprising program instructions, the computer program being loadable onto a data processing unit and adapted to cause the implementation of a detection method as defined above when the computer program is implemented on the data processing unit.
[0026] The invention also relates to a readable information medium on which a computer program product as defined above is stored.
[0027] Further provided is a computer embedded in a transport platform configured to implement the detection method as defined above, the electronic system being the on-board computer. The invention also relates to a transport platform comprising the computer as defined above.
[0028] The invention also relates to a method for generating a reference set, the reference set being representative of the operation of an electronic system in operation without anomalies, the electronic system comprising one or more processors each composed of a plurality of hardware blocks, the hardware blocks comprising execution hardware blocks and counting hardware blocks, the processor being adapted to execute at least one application by interactions of the application with said execution hardware blocks, each counting hardware block being adapted to measure an operating parameter representative of the interactions of the application with one of the execution hardware blocks to obtain measurements of said representative parameter, the generation method comprising at least the following steps:
[0029] - execution of the application on a test bench according to a plurality of predetermined input data, said input data being representative of the operation in operation of the electronic system without anomalies,
[0030] - measurements by at least one of the counting hardware blocks of at least one operating parameter representative of the interactions of one of the execution hardware blocks during the execution step with the application, to obtain measurements,
[0031] - processing of measurements of at least one operating parameter, to obtain processed data, and
[0032] - generation of the reference set from the processed data.
[0033] According to other embodiments, the method comprises one or more of the following characteristics, taken in isolation or in all technically possible combinations: during the measurement step, the measurements obtained are measurements of a plurality of operating parameters, in particular of a number of operating parameters between four and six, all of the measurements obtained being, for example, only the operating parameters measured by a portion of the hardware counting blocks.
[0034] - at least one new iteration of the steps of the generation method is implemented, the measurement step of one iteration being implemented by at least one hardware counting block different from the at least one hardware counting block implementing the measurement step of another iteration. each operating parameter representative of the interactions of one of the execution hardware blocks with the application is chosen from the group consisting of: the type of data processed by the hardware block considered, the type of instructions executed by the hardware block considered; the number of accesses to the hardware block considered, the number of calculations executed and / or the calculation time of the hardware block considered, the number of malfunctions of the hardware block considered, the number of interactions on an interconnection network internal to the processor involving the hardware block,and the number of interactions with the outside of the processor by the execution hardware block considered. each execution hardware block is chosen from the group consisting of: one or more calculation units, one or more branch prediction units, one or more memory registers, one or more cache memories, one or more random access memories, one or more processing chains, one or more memory protection or address translation units, one or more interconnection buses or networks, and one or more input-output units, each parameter representative of the operation of one of the execution hardware blocks being chosen from the group consisting of: the number of data of the same type processed by the calculation unit(s), the interconnection network(s) or the memory(s), the number of instructions of the same type executed by the calculation unit, the energy consumption of the calculation unit(s),the temperature of the computing unit(s), the number of successes and / or mispredictions of the branch prediction unit, the number of memory accesses, the number of successes and / or mispredictions of the cache memory, the number of successes and / or mispredictions of the memory protection unit, the execution time of an instruction in the processing chain, the number of information exchanges on the internal interconnection network(s) of the processor, the number of information exchanges via the interconnection network(s) per source, the number of information exchanges via the interconnection network(s) per destination, and the number of information exchanges via the input-output unit. the execution of the application is implemented according to at least two successive time phases,the processing step being implemented by applying respective operations to the measurements of each phase the application is configured to implement at least two functions, each function being implemented during a specific time phase.,
[0035] - each time phase corresponds to a predetermined time interval of execution of the application, in particular a time interval of less than 300 milliseconds. the processing step comprises an aggregation of the measurements for each operating parameter to obtain a database specific to each operating parameter. the set of measurements and the set of processed data occupy a respective size on a memory, the processing step comprising the application of an operation allowing the size of the set of processed data to be less than the size of the set of measurements. the operation is implemented by a machine learning technique comprising training from the set of measurements. the machine learning technique comprises training a neural network from the set of measurements.the operation is a statistical analysis operation such that the set of processed data is a statistical distribution of the set of measurements. a predetermined number of iterations of the measurement step is implemented in order to obtain a first set of measurements and a first set of processed data, the generation method further comprising the following steps: second implementation of the predetermined number of iterations of the measurement step to obtain a second set of measurements and a second set of processed data,.
[0036] - comparison of the first set of processed data and the second set of processed data according to a predetermined convergence criterion, when the convergence criterion is not met, implementation of a number of iterations of the measurement step greater than the predetermined number of iterations, and
[0037] - when the convergence criterion is met, generation of the reference set from the first and second sets of processed data.
[0038] The invention also relates to a reference game capable of being obtained by a generation method as defined above.
[0039] The invention also relates to the use of the reference set as defined above for detecting an anomaly in the electronic system in operation, in particular a computer attack against the electronic system, by comparing at least one measurement of at least one of the operating parameters during operation of the electronic system and by comparing said measurement with the reference set to detect possible inconsistencies between the measurement and said reference set.
[0040] The invention also relates to a computer program product comprising a readable information medium, on which is stored a computer program comprising program instructions, the computer program being loadable onto a data processing unit and adapted to cause the implementation of a generation method as defined above when the computer program is implemented on the data processing unit.
[0041] The invention also relates to a readable information medium on which is stored a computer program comprising program instructions, the computer program being loadable onto a data processing unit and adapted to cause the implementation of a generation method as defined above when the computer program is implemented on the data processing unit.
[0042] The invention will be better understood with the aid of the following description, given solely by way of non-limiting example and made with reference to the appended drawings, in which: [Fig 1] Figure 1 is a schematic representation of an electronic system according to the invention embedded in a transport system,
[0043] [Fig 2] Figure 2 is a flowchart of a method, according to the invention, for generating a reference set,
[0044] [Fig 3] Figure 3 is a flowchart of the slicing step of the generation process,
[0045] [Fig 4] Figure 4 is a flowchart of the measurement and processing steps of the generation process,
[0046] [Fig 5] Figure 5 is a flowchart of a method, according to the invention, for detecting an anomaly,
[0047] [Fig 6] Figure 6 is a schematic representation of a processor embedded in the electronic system of Figure 1.
[0048] Figure 1 represents an electronic system 10.
[0049] The electronic system 10 is embedded in a transport platform 12. The transport platform 12 is particularly suitable for transporting passengers and / or goods.
[0050] The transport platform 12 is, in particular, an avionics transport platform such as an airplane, a helicopter or a drone.
[0051] Alternatively, the transport platform 12 is a space transport platform such as a satellite or a space shuttle; a rail transport platform, in particular a train; a maritime transport platform such as a ship or a submarine or an automobile transport platform such as an autonomous car or a bus.
[0052] Alternatively, the invention applies to any field in which critical embedded computers are present. In particular, as an alternative, the electronic system 10 is embedded in a medical device, in a portable computer, in a home automation system or in an industrial system.
[0053] The platform 12 optionally comprises a plurality of electronic systems 10. For example, when the transport platform 12 is an aircraft, the platform comprises a take-off system, a piloting system, a communication system, an alert system, etc.
[0054] As seen in Figure 1, the electronic system 10 comprises one or more processors 14, an operating system 16 and a memory 18. The operating system 16 (often called OS from the English “Operating System”) is a set of programs which directs the use of the resources of the electronic system 10 by applications.
[0055] Each processor 14 is composed of a plurality of hardware blocks.
[0056] The hardware blocks include 20 execution hardware blocks and 22 counting hardware blocks.
[0057] A hardware execution block 20 is an electronic component configured to execute an elementary function of the processor 14 such as calculation, memory storage or data transfer.
[0058] In particular, the processor 14 is adapted to execute at least one application by interactions of the application with the execution hardware blocks 20.
[0059] The application is a program designed to perform a task, or a set of elementary tasks. The application runs using the services of the operating system to use the hardware resources provided by the hardware execution blocks 20.
[0060] In particular, each execution hardware block 20 is chosen from the group consisting of:
[0061] - one or more computing units,
[0062] - one or more branch prediction units,
[0063] - one or more internal memory registers of the processor 14,
[0064] - one or more cache memories, one or more RAM memories,
[0065] - one or more processing chains, one or more memory protection or address translation units, one or more interconnection buses or networks, and
[0066] - one or more input-output units.
[0067] These different hardware execution blocks 20 make it possible to provide the hardware resources necessary for the execution of the different applications.
[0068] Each counting hardware block 22 is adapted to measure an operating parameter representative of the interactions of the application with one of the execution hardware blocks 20 with the application to obtain measurements of said representative parameter.
[0069] Advantageously, the electronic system 10 comprises a plurality of hardware counting blocks 22, each capable of measuring a specific operating parameter. The electronic system 10 notably comprises between four and six hardware counting blocks 22.
[0070] Alternatively, the electronic system 10 comprises more than six hardware counting blocks 22.
[0071] Each representative parameter associated with one of the execution hardware blocks 20 is chosen from the group consisting of: the type of data processed by the execution hardware block 20; the type of instructions executed by the execution hardware block 20; the number of accesses to the execution hardware block 20;
[0072] - the number of calculations executed and / or the calculation time of the execution hardware block 20; the number of malfunctions of the execution hardware block 20, and the number of interactions on an interconnection network internal to the processor involving the execution hardware block 20, and the number of interactions with the exterior of the processor 14 by the execution hardware block 20.
[0073] Different data types are for example a boolean type, an integer type, a floating point real type, a character type, etc.
[0074] The different instructions are for example a calculation instruction, a data sending instruction, a storage instruction, etc.
[0075] Each internal interconnection network of the processor 14 forms a gateway allowing communication between different entities of the processor, in particular between the different hardware blocks.
[0076] More specifically, each parameter representative of the operation of one of the hardware execution blocks 20 is chosen from the group consisting of: the number of data of the same type processed by the calculation unit(s), the interconnection network(s) or the memory(s), the number of instructions of the same type executed by the calculation unit(s), the energy consumption of the calculation unit(s),
[0077] - the temperature of the calculation unit(s), the number of successes and / or errors in prediction of the branch prediction unit, the number of accesses to the memories, the number of successes and / or errors in requests to the cache memory, the number of successes and / or errors in requests to the memory protection unit, the execution time of an instruction in the processing chain, the number of exchanges of information via the interconnection network(s) internal to the processor 14, the number of exchanges of information via the internal interconnection network(s) obtained by a given source, the number of exchanges of information via the internal interconnection network(s) sent to a given destination, and the number of exchanges of information via the input-output unit.
[0078] Memory 18 stores a set of reference data, or also called a reference set hereinafter.
[0079] The reference data set is representative of the operation of the electronic system 10 without anomalies.
[0080] The term “anomaly-free operation” means an operation of the electronic system 10 as planned and envisaged during its design and operation. An abnormal operation is therefore an operation deviating from what was envisaged during the operation of the electronic system 10, due for example to a computer attack or to a hardware and / or software failure. In particular, an anomaly-free operating margin is defined around the operating values planned during the design of the system. For example, an anomaly-free operating temperature range of the computing units is determined at the design stage. An abnormal operation is detected when the temperature is, for example, 10°C warmer than the maximum limit of the anomaly-free operating range.
[0081] As will be explained later, the reference data set, or reference set, allows the detection of possible inconsistencies between the measurement of an operating parameter and the reference data set.
[0082] A method of generating 100 such a reference set will now be described, with reference to Figure 2 representing a flowchart of said generation method 100.
[0083] The generation method 100 comprises a step 120 of executing the application A on a test bench as a function of a plurality of predetermined input data E.
[0084] A test bench is understood to mean a test environment allowing the electronic system 10 to be placed in configurable and controlled conditions of use in order to observe and measure its behavior. Execution on a test bench is therefore different from the subsequent operational use of the electronic system 10.
[0085] The input data E are representative of the operational operation of the electronic system 10 without anomalies.
[0086] These input data E are determined according to the intended use for the electronic system 10 in operation. Feedback from the operation of similar electronic systems is also used. These input data are chosen in order to cover a wide variety of events likely to be encountered by the electronic system 10 during its normal operation, without anomalies.
[0087] Application A is configured to implement at least two functions, each function being implemented during a specific time phase P.
[0088] For example, application A includes an initial data acquisition phase, a data processing phase, a storage phase and a data sending phase.
[0089] Thus, the execution of application A is implemented according to at least two successive time phases P.
[0090] The generation method 100 therefore comprises a prior step 110 of dividing the application A into phases P.
[0091] Indeed, as will be explained in more detail later, global processing of an entire application whose behavior would vary significantly during the execution of the application would risk leading to the creation of an insufficiently selective reference set, which would make it difficult to detect abnormal behavior.
[0092] Depending on the type of application, the division is a temporal division or a functional division.
[0093] In particular, Figure 3 presents a decision tree for choosing the type of splitting of application A.
[0094] Thus, time division is chosen when the source code of application A is not known or mastered and the identification of phases is not easy.
[0095] Conversely, when the source code is controlled and the behavior of application A is not dependent on the input data, it is possible to identify phases P of application A and the functional breakdown will be favored.
[0096] Indeed, performing a functional breakdown and identifying the behavioral P phases of an application requires a certain mastery of the application, which is not always the case when integrating third-party applications. In this "black box" context, classic profiling solutions (such as using the well-known Gprof software) or static code analysis can help identify the application phases, and enable such a functional breakdown. This profiling searches in particular for regular loops, i.e. sets of instructions whose conditionals and loop bounds can be expressed in the form of affine functions of the enclosing iterators.
[0097] However, if the behavior of application A is highly data-dependent, as may be the case for example for cryptography applications, a behavioral functional breakdown is not possible, as it cannot be repeated from one execution to another.
[0098] In this case, the slicing is a time slicing which requires an additional step of selecting the sampling frequency.
[0099] This choice has an impact on the total number of P phases and is determined by seeking a compromise between the finesse of the characterization of these phases and the size of the reference set obtained.
[0100] For example, in periodic embedded applications, the division is carried out on the basis of this period.
[0101] However, if the activation of a task itself is composed of phases such as a data retrieval phase, a calculation phase and a result delivery phase, the sampling frequency is more important.
[0102] In particular, each time phase P corresponds to a predetermined time interval of execution of the application, in particular a time interval of less than 300 milliseconds, typically 200 milliseconds.
[0103] The generation method 100 comprises a measurement step 130 by at least one of the counting hardware blocks 22 of at least one operating parameter representative of the interactions of one of the execution hardware blocks 20 during the execution step with the application, to obtain measurements. The measurements obtained are therefore measurements of a plurality of operating parameters as defined above.
[0104] Figure 4 illustrates this measurement step 130 in particular.
[0105] These measurements are carried out phase by phase in order to obtain a set of measurements of the operating parameters per phase.
[0106] At least one new iteration I of the steps of the generation method 100 is implemented, the measurement step 130 of an iteration being implemented by at least one hardware counting block 22 different from the at least one hardware counting block 22 implementing the measurement step of another iteration I. A set of hardware counting blocks 22 measuring the operating parameters during an iteration I is called configuration C of the hardware counting blocks 22.
[0107] As explained above, it is not possible to determine a priori which hardware events will be targeted and therefore requested by possible future computer attacks, and having only a reduced number of hardware counting blocks 22 compared to the large number of events, it is therefore interesting to fairly browse all the hardware events and to vary the configurations of the hardware counting blocks 22.
[0108] In order to collect statistically significant data, application A is executed over a large number of iterations I, with different input data, to capture distributions of the possible values of the different operating parameters under nominal operation.
[0109] The generation method 100 then comprises a step 140 of processing the measurements of the operating parameters, to obtain processed data.
[0110] In particular, the processing step 140 is implemented by applying respective operations to the measurements of each phase.
[0111] The processing step 140 notably comprises an aggregation of the measurements for each operating parameter to obtain a database specific to each operating parameter.
[0112] In particular, the processing 140 comprises an aggregation of the measurements by phase P, measurements by configuration C, and measurements by iteration I to obtain a set of the collected statistical data. At the end of each of these steps, the collected values are aggregated as follows.
[0113] During a phase, all the pairs (event, value) are collected such that the event belongs to the current configuration of counting hardware blocks 22.
[0114] Aggregation by phase P consists of grouping all the measured values in a table indexed by phase.
[0115] Aggregation by configuration C extends the set of events beyond the current configuration to cover the set of configurations necessary to cover all hardware events expected during normal operation.
[0116] Iteration aggregation I differs from the previous aggregation by replacing the (event, value) pairs with (event, set of values observed during the different iterations) pairs. The set of measurements collected for application A is thus likely to be very large and typically represent several gigabytes of collected data.
[0117] It is then useful to reduce this data to obtain a reference set J of reduced size, which can be integrated into the memory 18 of the electronic system 10.
[0118] The processing step 140 then comprises the application of an operation allowing the memory size of all the processed data to be less than the memory size of the set of measurements.
[0119] In particular, the operation is implemented by a machine learning technique including training from the measurement set.
[0120] Machine learning is implemented by training a neural network based on the measurement set. The reference set would then become this network once trained. Symbolic learning or segmentation techniques are used as alternatives.
[0121] Alternatively, the reduction operation is a statistical analysis operation such that the processed data set is a statistical distribution of the measurement set.
[0122] This operation is similar to that commonly performed to enable the visualization of statistical data with histograms or box plots for example.
[0123] The statistical distribution includes a plurality of values giving key information about the statistical distribution such as the minimum, maximum, mean, median, standard deviation, deciles, quartiles, etc. For example, these values are the different intervals ("bins" in English) of a histogram or the quartiles for box plots.
[0124] This plurality of values then becomes the reference set J.
[0125] It is therefore necessary to reduce the size of the collected data, but it is also important to ensure that the collected data is statistically representative of the application. It is therefore necessary to evaluate the statistical coverage of the collected data.
[0126] Thus, a predetermined number of iterations of the measurement step 130 is implemented in order to obtain a first set of measurements and a first set of processed data.
[0127] The generation method 100 then further comprises a second implementation of the predetermined number of iterations of the measurement step 130 to obtain a second set of measurements and a second set of processed data. Then, the first set of processed data is compared to the second set of processed data according to a predetermined convergence criterion.
[0128] The convergence criterion determines whether these two distributions are statistically similar. The literature proposes different solutions to this problem, such as comparing histograms by calculating the Bhattacharyya distance, the Kullback-Leiber divergence, or the Hellinger distance.
[0129] When the convergence criterion is not met, the generation method 100 comprises implementing a number of iterations of the measurement step 130 greater than the predetermined number of iterations.
[0130] In other words, the two collected statistical distributions are concatenated and the collection of a distribution of larger size, for example of double size, is repeated.
[0131] This step is repeated until the statistical distribution converges.
[0132] When the convergence criterion is met, the generation method 100 comprises the generation of the reference set J from the first and second sets of processed data.
[0133] A reference set J is thus obtained from the processed data. As will be explained below, this reference set is notably used to detect an anomaly in the electronic system 10 in operation, in particular a computer attack against the electronic system 10, by comparing at least one measurement of at least one of the operating parameters during the operation of the electronic system 10 and by comparing said measurement with the reference set to detect possible inconsistencies between the measurement and the reference set.
[0134] However, other applications of the reference set J thus generated are possible.
[0135] Indeed, in critical fields such as automotive, railway, avionics or space, the current trend in embedded electronic / computer systems is to move towards the use of generic computing platforms ensuring both functionality and operational safety and security properties.
[0136] These computing platforms have a limited set of hardware resources and must run one or more application software, of different criticality levels and from different suppliers.
[0137] The complexity of new embedded systems for critical areas and the market sharing between different economic players has required the modification of the industrial process by redefining the roles and responsibilities between the different stakeholders during the design of a system. Each of these players is responsible for providing the technological building blocks assigned to them and for applying the constraints imposed on them by the operating requirements of the system.
[0138] The hardware platform supplier delivers the calculator as well as all the electronics necessary for the proper functioning of the software applications that may be assigned to it.
[0139] The operating system provider provides the operating system that will manage the execution of the various application software according to their needs in terms of priority, execution time and frequency, memory space, etc.
[0140] The application software provider is responsible for the proper functioning of its applications while respecting the instructions provided by the system integrator regarding the rules for using the system's hardware resources.
[0141] Finally, the system integrator is entrusted with the central role. He assembles the various technological building blocks, both hardware and software, of the system while allocating all the necessary resources to the software application suppliers. In addition, he is responsible for consolidating the operating requirements and taking into account the interactions of the various technological building blocks with respect to the system resources. He is the one who must guarantee the overall operational safety and security of the system.
[0142] Control of the system involves all participants, each at their own level of responsibility, but it is the system integrator who is responsible for defining and validating the entire system being built, ensuring that the applications running on the system's computing platforms comply at all times with non-functional specifications such as time, energy and sharing of resources required for computing, safety, security, etc.
[0143] In this context, the reference set J is advantageously used to transmit the requirements of the system integrator to the other actors and the transmission of non-functional properties of the applications to the integrator such as the actual need in terms of shared hardware resources.
[0144] Indeed, by providing the corresponding reference set with its application, the software provider provides information linked to the use of hardware resources by its application, and therefore a rate of use of the different shared resources.
[0145] The J reference set is also a way for the system integrator to translate his requirements into a maximum resource usage envelope for each application, so as to guarantee their coexistence in the final integrated system.
[0146] The reference set J representing the hardware behavior of applications therefore has several areas of applicability.
[0147] The J reference set can be used in a cybersecurity framework, as will be explained later, to protect against external cyber attacks.
[0148] In a safety framework, the reference set J can also be used to participate in system monitoring, detect abnormal cases linked to failures and trigger measures aimed at returning to a stable state.
[0149] The J reference set can also be used as a means of transmitting non-functional requirements. Indeed, in the context of multi-core processors, themselves complex systems from third-party suppliers and most often having partial specifications, the usual technique of time partitioning has become insufficient.
[0150] The J benchmark set can also be used to facilitate the integration of applications by taking into account their specific use of shared hardware resources. The usual technique based on abstract models is no longer suitable for multi-core applications.
[0151] The reference set J can also be used to build a library of applications representative of a domain by characterizing them via the distance between their respective reference sets.
[0152] In an incremental certification framework, the reference set J can also be used to anticipate the impact of integrating a new application into the system knowing its reference set.
[0153] A method 200 for detecting an anomaly in an electronic system 10 in operation will now be described with reference to Figure 5 representing a flowchart of said detection program 200.
[0154] The detection method 200 is carried out using a reference set as generated above.
[0155] In particular, the method 200 aims to detect a computer attack against the electronic system 10.
[0156] However, the detection method 200 is also capable of detecting a hardware and / or software failure of the electronic system 10.
[0157] Several embodiments are possible for implementing the detection method, as illustrated in Figure 6. According to a first embodiment, the detection method is implemented by a dedicated programmable logic circuit forming one of said hardware execution blocks 20 of the processor 14.
[0158] Thus, this implementation is carried out at the hardware level in the form of a dedicated component, as shown schematically in Figure 6 with the reference numeral 1. This component could thus reprogram and operate the hardware counting blocks 22 without worrying about the privilege levels required at the software level. A hardware component also makes it easy to have a dedicated memory for storing the reference sets within the component itself.
[0159] This embodiment therefore offers a highly secure solution.
[0160] According to a second embodiment, the detection method is implemented by software of an operating system 16 of the electronic system 10 configured to interact with a memory protected by a memory protection unit.
[0161] This embodiment is shown schematically in Figure 6 with the reference numeral 2.
[0162] In this case, the reference set is stored in main memory. The software will benefit from the high privilege levels of the operating system to configure and read the hardware counting blocks 22. The storage area of the reference sets is protected via the memory management unit (MMU).
[0163] According to a third embodiment, the detection method is implemented by an application suitable for being implemented by the electronic system 10 by adding a driver controlling the hardware blocks, as shown schematically in Figure 6 with the numerical reference 3.
[0164] Thus, the method is implemented in the form of an application running at the same level as the applications to be monitored.
[0165] This embodiment is the easiest to implement.
[0166] Alternatively, the detection method is implemented via a hybrid implementation, both software and hardware, such as a software-level implementation that stores the reference sets in a dedicated hardware flash memory, to protect the memory space of the reference sets.
[0167] In the initial state of the detection method 200, the electronic system 10 is in operational operation. For example, the electronic system 10 is on board an aircraft flying to an airport. The electronic system 10 is then, for example, an avionics system for communication, piloting, etc.
[0168] The detection method 200 comprises a first measurement step 210, during operation of the electronic system 10, of at least one parameter representative of the interactions of the application A with one of the execution hardware blocks 20 to obtain measurements of said representative parameter.
[0169] The representative parameters are as defined above. They correspond in particular to those for which measurements were carried out during the process of generating the associated reference set.
[0170] For each measurement, the detection method 200 comprises the comparison 220 of the measurement with the associated reference data set, or reference set, to detect possible inconsistencies between the measurement and the reference data set.
[0171] An inconsistency is a statistical deviation of the measurement from the reference set. For example, the measurement is not included in the envelope formed by the reference set.
[0172] Any deviation indicates abnormal application behavior that could be related to a cyber attack.
[0173] Thus, the detection method 200 comprises a step 230 of issuing an alert W according to an alert criterion relating to the inconsistency(ies) detected.
[0174] Each alert criterion depends on at least one condition on the representative parameters. Each alert criterion is, for example, a threshold value of the measured parameter.
[0175] During the generation phase 100 of the reference set J, it is possible to build an exhaustive reference set, at the cost of many executions of the application, with a different set of events. On the contrary, during the detection phase 200, it is not possible to execute the application several times in order to detect an attack. Detection must take place directly, during execution.
[0176] Thus, only a limited number of parameters are measured from 22 hardware counting blocks in order to minimize the detection time.
[0177] According to one embodiment, the measured parameters are alternated rapidly in order to cover all significant events by sampling in a short period of time.
[0178] Alternatively, the measured parameters are chosen by means of a hierarchical selection of the monitored events. When at least one alert W associated with one of the execution hardware blocks 20 is issued, a new iteration of the steps of the detection method 200 starting from the measurement step 210 is implemented.
[0179] The set of execution hardware blocks 20 for which a representative parameter measurement is carried out being, at each iteration, included in the set of execution hardware blocks 20 of the previous iteration, the set of execution hardware blocks 20 being in particular included in the set of execution hardware blocks 20 associated with the alert(s).
[0180] Thus, a small number of events is chosen, corresponding to the number of available counting hardware blocks 22 and covering the different execution hardware blocks 20 of the processor 14, in order to constitute a first level of monitoring.
[0181] If one or more of these events exceed thresholds from the reference set, a suspicion of attack is raised and the measured parameters are reconfigured to focus on the relevant execution hardware block 20 in order to confirm or deny the suspicion.
[0182] The measurement steps 210 are thus advantageously repeated on several hierarchical levels. This method makes it possible to improve the reliability of the detection despite the limited number of hardware counting blocks 22.
[0183] The detection method 200 comprises, when an alert W associated with one of the execution hardware blocks 20 is issued, a new iteration of the steps of the detection method 200 starting from the measurement step 210, the alert criterion comprising, at each iteration, an increasing number of conditions.
[0184] This makes it possible to refine the alert and characterize the suspicion of an attack more precisely.
[0185] For example, a first alert condition is the exceeding of a memory access threshold of a hardware block. At the next iteration, the alert condition includes, in addition to this threshold, the exceeding of a threshold of calculations executed by this hardware block.
[0186] The method 200 then comprises analyzing 240 the nature of the execution hardware block 20 associated with each inconsistency, analyzing the temporal frequency of the inconsistencies, and / or analyzing the deviation of each measurement associated with the inconsistencies from the reference data set.
[0187] The detection method 200 then comprises a detection step 250, depending on the or each analysis of an anomaly, in particular a computer attack or a hardware failure of one of the hardware blocks. Indeed, depending on the frequency, type and severity of the abnormal events, it is possible to distinguish a computer attack from a hardware failure. In particular, the transient events linked to operational safety are essentially punctual, and result at the level of the counting hardware blocks 22 by a peak.
[0188] Cybersecurity events, on the other hand, last as long as the attack, and attacks generally have a longer-lasting effect on hardware counting blocks 22.
[0189] By comparing the inconsistency associated with the issued alert with a set of predetermined known anomalies, the method comprises categorizing the alert when the or each inconsistency matches one of the known anomalies of said set of anomalies.
[0190] It is thus possible to know the nature of the anomaly encountered, in particular the type of computer attack, and to react accordingly.
[0191] However, the detection method 200 does not depend on the type of attack. The method also makes it possible to detect new classes of attacks targeting hardware or software.
[0192] Finally, as components age, the frequency of transient events gradually increases until permanent failure occurs. The detection process, via this analysis, makes it possible to observe this frequency variation, to participate in the preventive maintenance of components and prevent this type of failure.
Claims
26 DEMANDS 1. A method for detecting (200) an anomaly in an operating electronic system (10), in particular a computer attack against the electronic system (10), the electronic system (10) comprising one or more processors (14) each composed of a plurality of hardware blocks (20, 22), the processor (14) being adapted to execute at least one application (A) through interactions of the application (A) with said hardware blocks (20), the detection method (200) comprising at least the following steps: measurement (210), during the operation of the electronic system (10), of at least one parameter representative of the interactions of the application (A) with one of the hardware blocks (20) to obtain measurements of said representative parameter, - for each measurement, comparison (220) of said measurement with an associated reference data set (J) to detect any inconsistencies between the measurement and the reference data set (J), said reference data set (J) being representative of the operation of the electronic system (10) without anomalies, and issuance (230) of an alert (W) according to an alert criterion relating to the inconsistencies detected.
2. A detection method (200) according to claim 1, wherein the detection method (200) further comprises at least one step selected from the group consisting of: - analysis (240) of the nature of the material block (20) associated with each inconsistency, - analysis (240) of the temporal frequency of inconsistencies, and analysis (240) of the deviation of each measurement associated with inconsistencies from the reference dataset (J); the detection method (200) comprising a detection step (250), based on the analysis of an anomaly, in particular a computer attack or a hardware failure of one of the hardware blocks (20).
3. A detection method (200) according to claim 1 or 2, wherein when at least one alert (W) associated with one of the hardware blocks (20) is issued, a new iteration of the steps of the detection process (200) from the measurement step (210) is implemented, the set of hardware blocks (20) of which a representative parameter measurement is carried out being, at each iteration, included in the set of hardware blocks (20) of the previous iteration, the set of hardware blocks (20) being in particular included in the set of hardware blocks (20) associated with the alert(s) (W).
4. Detection method (200) according to any one of claims 1 to 3, wherein each alert criterion depends on at least one condition on the representative parameters, the detection method (200) comprising, when an alert (W) associated with one of the hardware blocks (20) is issued, a new iteration of the steps of the detection method (200) from the measurement step (210), the alert criterion comprising, at each iteration, an increasing number of conditions.
5. A detection method (200) according to any one of claims 1 to 4, wherein each representative parameter associated with one of the material blocks (20) is chosen from the group consisting of: - the type of data processed by the hardware block (20); the type of instructions executed by the hardware block (20); the number of accesses to the hardware block (20); - the number of calculations executed and / or the calculation time of the hardware block (20); the number of malfunctions of the hardware block (20), and the number of interactions on an internal processor interconnection network (14) involving the hardware block (20), and the number of interactions with the outside of the processor by the hardware block (20).
6. A detection method (200) according to any one of claims 1 to 5, wherein each material block (20) is selected from the group consisting of: - one or more computing units, - one or more branch prediction units, - one or more internal memory registers of the processor, - one or more cache memories, one or more RAM memories, - one or more processing chains, one or more memory protection or address translation units, one or more buses or interconnection networks, - one or more input / output units, each parameter representing the operation of one of the hardware blocks (20) being chosen from the group consisting of: the number of data of the same type processed by the processing unit(s), the interconnection network(s) or the memory(ies), the number of instructions of the same type executed by the processing unit(s), the energy consumption of the processing unit(s), - the temperature of the processing unit(s), the number of successes and / or errors in branch prediction, the number of memory accesses, the number of successes and / or errors in cache memory requests, the number of successes and / or errors in memory protection unit requests, the execution time of an instruction in the processing chain, the number of information exchanges via the internal processor interconnection network(s), the number of information exchanges via the interconnection network(s) obtained from a given source, the number of information exchanges via the interconnection network(s) sent to a given destination, and the number of information exchanges via the input / output unit.
7. Detection method (200) according to any one of claims 1 to 6, wherein the electronic system (10) comprises a plurality of counting material blocks (22) each configured to measure one of the representative parameters, the measurement step (210) comprising the measurement of a plurality of representative parameters, in particular between four and six, for example the representative parameters measured by a part of the counting material blocks (22).
8. A detection method (200) according to any one of claims 1 to 7, wherein the detection method (200) further comprises a step of:
29. Comparison of the inconsistency(ies) associated with the alert (W) issued with a set of predetermined known anomalies, and - categorization of the alert (W) when the inconsistency or each inconsistency corresponds to one of the known anomalies of said set of anomalies.
9. Detection method (200) according to any one of claims 1 to 8, wherein the electronic system (10) is a computer embedded in a transport platform (12).
10. Detection method (200) according to any one of claims 1 to 9, wherein the detection method (200) is implemented by a system selected from the group consisting of: a dedicated programmable logic circuit forming one of said hardware blocks (20) of the processor; software of an operating system (16) of the electronic system (10) configured to interact with a memory protected by a memory protection unit, and an application suitable for implementation by the electronic system (10) by adding a driver controlling the hardware blocks (20).
11. Product computer program comprising a readable information carrier, on which is stored a computer program comprising program instructions, the computer program being loadable onto a data processing unit and adapted to drive the implementation of a detection method (200) according to any one of claims 1 to 10 when the computer program is implemented on the data processing unit.
12. Readable information medium on which a computer program product according to claim 11 is stored.
13. Computer embedded in a transport platform (12) configured to implement the detection method (200) according to any one of claims 1 to 10, the electronic system being the computer embedded.
14. Transport platform (12) comprising the computer according to claim 13.
