Communication relay device and communication relay method
By introducing feature value extraction, frequency determination and phase determination components into the communication relay device, differentiating and reducing replica communication in DoS attacks, the problem of missed formal communication in the prior art is solved, and the replica communication is realized Effective defense and control the stability of the system.
Patent Information
- Application Number
- CN202180046312.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Priority Date
- 2020-08-18
- Filing Date
- 2021-06-15
- Publication Date
- 2024-12-06
- Estimated Expiration
- 2041-06-15
AI Technical Summary
The prior art is difficult to distinguish and effectively defend against DoS attacks in the form of formal communication content that are copied and sent by large quantities, resulting in formal communication being cut off and the target device cannot be continued to be controlled.
A communication relay device is designed to extract the characteristic values of communication data through the characteristic value extraction unit, and to calculate whether the frequency of the characteristic value and the reception timing match the predetermined period by using the frequency determination unit and the phase determination unit, respectively, and output a cut-off signal. and phase matching signals, based on these signals, improper communication data is discarded.
It effectively defends against DoS attacks that copy a large number of formal communication content, avoids the situation where formal communication is cut off, and ensures the stable operation of the control target equipment.
Smart Images

Figure CN115836509B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to a communication relay device and a communication relay method. Background Art
[0002] Control systems that control equipment such as chemical plants and product assembly equipment in the manufacturing industry need to maintain safe and stable operation. The control systems are controlled in real time based on communication content regularly and periodically received from the outside via a network used for such control systems.
[0003] However, there are cases where data with improper contents is sent to the control system via the control system network with the intention of hindering the operation of the control system or causing it to erroneously operate. As a countermeasure, there is a method known to connect the device to be protected to the control system network via a firewall or IPS (Intrusion Prevention System) that has the ability to cut off such improper communication.
[0004] For example, a DoS (Denial of Services) attack is known as an attack on a control system, in which a large amount of data is sent to a server as a target of the attack. If a DoS attack occurs, the server to which the data is sent may fail due to the load.
[0005] As a countermeasure against such DoS attacks, Patent Document 1 discloses a network system that extracts feature values from communication content, calculates arrival frequency for each feature value, and cuts off communication having the feature value when the arrival frequency is high, thereby avoiding DoS attacks.
[0006] Prior art literature
[0007] Patent Literature
[0008] Patent Document 1: Japanese Patent Application Publication No. 2004-318742 Summary of the invention
[0009] However, in the technology disclosed in Patent Document 1, it is difficult to avoid DoS attacks in the form of monitoring regular communications and copying and sending the communication content in large quantities. This is because, in the prior art, if the regular communication content is copied, the characteristic value of the attack communication also becomes the same as the regular communication content, and it is impossible to distinguish. Therefore, if the arrival frequency of each characteristic value of the communication content is based on the communication content, while the improper communication of the DoS attack is cut off, the regular communication that should have passed is also cut off like the improper communication, resulting in the undesirable situation that the target device cannot be controlled any more.
[0010] The present invention has been made in view of the above-mentioned problems, and an object of the present invention is to provide a technology capable of effectively defending a protection target device from a DoS attack in which a large amount of legitimate communication contents are copied and sent.
[0011] In order to achieve the above-mentioned purpose, the communication relay device of the present invention relays the communication data received via the communication network to the protected object device, and the communication relay device is characterized in that it comprises: a feature value extraction unit that extracts feature values from the communication data; a frequency determination unit that calculates the average value of the frequency of each feature value extracted by the feature value extraction unit, and outputs a cut-off signal when the average value of the frequency of each feature value is above a threshold value of each feature value; a phase determination unit that determines whether the reception timing of the communication data matches the range of the phase corresponding to a specified period for each feature value extracted by the feature value extraction unit, and outputs a phase matching signal when it matches; and a discarding unit that discards the communication data based on the cut-off signal output by the frequency determination unit and the phase matching signal output by the phase determination unit.
[0012] Effects of the Invention
[0013] According to the present invention, it is possible to effectively defend a protection target device from a DoS attack in which a large amount of legitimate communication contents are copied and sent. BRIEF DESCRIPTION OF THE DRAWINGS
[0014] Figure 1 This is a block diagram showing the configuration of the communication relay device according to the first embodiment.
[0015] Figure 2 This is a time chart showing the time change of the signal in the average frequency determination circuit.
[0016] Figure 3 This is a block diagram showing the configuration of a communication phase determination circuit according to the first embodiment.
[0017] Figure 4 This is a timing chart showing an example of changes in time series of each signal inside the communication phase determination circuit according to the first embodiment.
[0018] Figure 5 This is a diagram showing the relationship between the output of each LPF and the phase quadrant determination result in the communication phase determination circuit of the first embodiment.
[0019] Figure 6 This is a diagram showing the relationship between each demodulated signal and the phase matching signal corresponding to each quadrant in the first embodiment.
[0020] Figure 7 It is a block diagram showing the configuration of a communication phase determination circuit according to the second embodiment.
[0021] Figure 8 This is a diagram showing the relationship between the output of each LPF and the phase quadrant determination result in the communication phase determination circuit according to the second embodiment.
[0022] Fig. 9 It is a diagram showing the relationship between each demodulated signal and the phase matching signal corresponding to each quadrant according to the second embodiment. DETAILED DESCRIPTION
[0023] Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. In addition, the embodiments described below are not intended to limit the invention described in the scope of the claims. In addition, not all of the elements and combinations thereof described in the embodiments are necessary for the solution of the invention. In addition, there are cases where illustrations and descriptions are omitted for structures that are necessary for the structure of the invention but are well-known. In addition, the number of elements shown in each figure is an example and is not limited to the illustrated case.
[0024] (First Embodiment)
[0025] First, the communication relay device 1 according to the first embodiment will be described. Figure 1 It is a block diagram showing the configuration of the communication relay device 1 according to the first embodiment.
[0026] The communication relay device 1 connects its input terminal Tin to the network 3 and connects its output terminal Tout to the protected device 2. The protected device 2 is a control device directly connected to a control object (not shown) such as a PLC (Programmable Logic Controller), and performs control actions (sensor drive) with respect to the control object at a predetermined control cycle. The control object is a chemical plant, assembly equipment, manufacturing equipment, etc.
[0027] The control device 4 is, for example, an industrial PC (Personal Computer), and instructs the protected device 2 on the content of processing. Therefore, the control device 4 periodically communicates with the protected device 2 via the network 3 based on the control cycle.
[0028] There is a concern that the attacker may also connect an attack device 9 to the network 3. The attack device 9 may monitor the communication data transmitted on the network 3 and perform a DoS attack by sending a large number of copies of the communication data to the protected device 2. If no countermeasures are taken against the DoS attack, the protected device 2 may not be able to process the large amount of communication data received, which may cause system failure.
[0029] Next, the internal structure of the communication relay device 1 will be described. The communication relay device 1 includes a feature value extraction circuit 13 , a plurality of average frequency determination circuits 15 , a plurality of communication phase determination circuits 17 , and a discarding circuit 19 .
[0030] The communication data 11 input from the input terminal Tin is input to the feature value extraction circuit 13 and the discarding circuit 19. In the present embodiment, the communication data 11 is a control frame or the like that indicates the processing content of the control object for the protection object device 2 to the protection object device 2, but is not limited to this. The feature value extraction circuit 13 analyzes the input communication data 11 and extracts and outputs its feature value 14. The feature value 14 may be a value that only extracts information (for example, a transmission source address) at a determined position within the communication data 11, or may be a value calculated from the entire content of the communication data 11 using a hash function or the like. The same feature value 14 is extracted from the same communication data 11, and different feature values 14 are extracted from different communication data 11. The feature value 14 extracted by the feature value extraction circuit 13 is input to a plurality of average frequency determination circuits 15 and a plurality of communication phase determination circuits 17.
[0031] The plurality of average frequency determination circuits 15 correspond to different characteristic values, respectively, and calculate the average frequency of the arrival frequency for each characteristic value 14 assigned, and determine whether the average frequency is greater than a predetermined threshold value. The characteristic value set for the average frequency determination circuit 15 can be a characteristic value of regular communication data set in advance, or can be obtained by the characteristic value extraction circuit 13 from regular communication data in a normal state that is not subjected to DoS attacks. The normal state refers to, for example, a situation where the reception frequency of communication data is lower than a predetermined value.
[0032] Each average frequency determination circuit 15 outputs a cutoff signal 16 indicating "high frequency" when the average frequency of the arrival frequency of each corresponding characteristic value is greater than a predetermined threshold value for each characteristic value. The average frequency can be calculated by counting the number of times received within a certain period of time as the average frequency, or by forming a low-pass filter based on a first-order lag system or the like and taking the magnitude of the output as the average frequency.
[0033] In addition, each average frequency determination circuit 15 can also calculate the specified threshold value of each characteristic value in advance based on the arrival frequency of each characteristic value of the communication data under normal conditions that are not subject to DoS attacks using the same method as the average frequency. For example, the result obtained by counting the number of reception times of the communication data of each characteristic value received within a certain period of time under normal conditions can be calculated as the specified threshold value of each characteristic value. Alternatively, the size of the output when the reception timing of the communication data of each characteristic value received within a certain period of time under normal conditions is used as the pulse input of the low-pass filter of the first-order lag system can be calculated as the specified threshold value of each characteristic value.
[0034] use Figure 2 The operation of each average frequency determination circuit 15 will be described in detail. Figure 2 1 is a time chart showing the time variation of the signal in the average frequency determination circuit 15. Figure 2 In the example, the characteristic value 14 is classified into three types: A, B, and C, and they arrive at each of them at a certain period under normal conditions. Figure 2 During the period shown as the DoS state, the communication data of the characteristic value A is copied by the attacker and is continuously sent in large quantities throughout the entire period of the DoS state.
[0035] Based on the above, the average frequency determination circuit 15 calculates the average arrival frequency for each characteristic value by using a low-pass filter of a first-order lag system that uses the reception timing of the communication data as a pulse input. Figure 2 In the normal state, the calculated values of the arrival frequencies of the characteristic values A, B, and C are all suppressed to low values below the predetermined threshold. However, in the DoS state, the calculated result of the arrival frequency of the characteristic value A becomes a high value above the predetermined threshold.
[0036] The average frequency determination circuit 15 determines whether the arrival frequency of each characteristic value calculated in this way exceeds a predetermined threshold value, and outputs a cutoff signal 16 for instructing the discarding circuit 19 to cut off the characteristic value if it exceeds (the frequency is large), and does not output the cutoff signal 16 if it does not exceed (the frequency is small). As a result, the discarding circuit 19 discards the communication data with a high frequency having the characteristic value A. In this case, since the cutoff signal 16 is not output for the communication data with the characteristic values B and C, the discarding circuit 19 does not cut off the communication data.
[0037] exist Figure 2In the case of the communication data of the characteristic value A, the cut-off signal 16 is output during the period from time t1 to t2, so that all the communication data of the characteristic value A input during this period are discarded. In this way, the regular communication data with the characteristic value A that arrives in the DoS state is cut off by only the cut-off signal 16. The communication phase determination circuit 17 described below improves this.
[0038] The communication phase determination circuit 17 determines whether each characteristic value 14 is a phase corresponding to a predetermined cycle, and outputs a phase matching signal 18 when the phase is within the range corresponding to the predetermined cycle (phase matching), and does not output a phase matching signal 18 when the phase is outside the range (non-phase matching). Figure 3 as well as Figure 4 A method for determining a phase in a predetermined cycle will be described based on this.
[0039] The discarding circuit 19 performs a discrimination operation of directly outputting the communication data 11 input from the input terminal Tin from the output terminal Tout or discarding the communication data 11 input from the input terminal Tin without outputting it, based on the presence or absence of the output of the cutoff signal 16 from the average frequency determination circuit 15 and the presence or absence of the output of the phase matching signal 18 from the communication phase determination circuit 17. Specifically, the discarding circuit 19 discards the input communication data 11 when the determination result of the average frequency determination circuit 15 is "high frequency" (the cutoff signal 16 is output).
[0040] However, even when the average frequency determination circuit 15 determines that the frequency is high, the discarding circuit 19 does not discard the input communication data 11 but directly outputs it when the communication phase determination circuit 17 determines that the phase matches (outputs the phase matching signal 18).
[0041] In this embodiment, the communication relay device 1 includes a plurality of average frequency determination circuits 15 and a plurality of communication phase determination circuits 17 corresponding to a plurality of characteristic values. However, the present invention is not limited thereto, and one average frequency determination circuit 15 and one communication phase determination circuit 17 may process a plurality of characteristic values respectively.
[0042] According to the above configuration, when the communication data 11 input from the input terminal Tin is high frequency, i.e., it is subjected to DoS attack, the communication relay device 1 discards the communication data 11, and directly outputs the communication data 11 whose receiving timing matches the phase corresponding to the predetermined period, i.e., it can be estimated as regular communication without discarding it. Therefore, it is possible to effectively defend the protection target device 2 from DoS attack in the form of massively copying and sending the content of legitimate communication.
[0043] The DoS attack communication data arriving at the timing of outputting the phase matching signal 18 will also pass through without being distinguished from the regular communication data. However, since the amount of DoS attack communication data reaching the protected device 2 can be reduced, system failure can be prevented.
[0044] Next, use Figure 3 as well as Figure 4 The internal structure and operation of the communication phase determination circuit 17 will be described. Figure 3 1 is a block diagram showing the configuration of the communication phase determination circuit 17 according to the first embodiment. Figure 4 This is a timing chart showing an example of changes in the respective signals inside the communication phase determination circuit 17 according to the first embodiment in accordance with the time series.
[0045] The communication phase determination circuit 17 includes a pulse signal generating circuit 170 , demodulation signal generating circuits 171 and 172 , demodulation circuits 173 and 174 , low-pass filters 175 and 176 , and a phase determination circuit 177 .
[0046] The pulse signal generating circuit 170 outputs a short pulse signal 1700 when the input characteristic value 14 matches the value set in the circuit itself. Figure 4 The example shown shows the waveform of the pulse signal 1700 in the case where regular communication data is input at a predetermined cycle in the normal state and communication data is input at a high frequency in the DoS state.
[0047] The demodulation signal generating circuits 171 and 172 generate demodulation signals 1710 and 1720 required for the synchronous demodulation process described below. The demodulation signals 1710 and 1720 generated here are rectangular wave signals with the same frequency as the control process in the control device 4, a predetermined period, and a magnitude of +1 or -1. Then, the demodulation signal 1710 generated by the demodulation signal generating circuit 171 and the demodulation signal 1720 generated by the demodulation signal generating circuit 172 are as shown in FIG. Figure 5 as well as Figure 6 The phases shown are shifted by 90 degrees.
[0048] The demodulation circuits 173 and 174 multiply the input pulse signal 1700 and the demodulation signals 1710 and 1720, respectively, and output them as demodulation outputs 1730 and 1740. However, even if multiplied, the demodulation signals 1710 and 1720 can only be +1 or -1. Figure 4 As shown, the demodulation outputs 1730 and 1740 maintain the sign of the pulse signal 1700 or invert the sign. Therefore, the demodulation circuits 173 and 174 do not need to use a general-purpose multiplication circuit, and only need to have a function of sign inversion.
[0049] According to the operation of the demodulation circuit, the demodulated outputs 1730 and 1740 based on the communication data arriving at the same frequency under the control process of the control device 4 have the same positive and negative signs every time. On the other hand, the communication data arriving at different frequencies, such as the communication data based on DoS attack, have uneven positive and negative signs, and if the averaging described below is performed, it will be almost zero and can be removed.
[0050] Low pass filters (LPF) 175 and 176 have, for example, input-output characteristics of a first-order lag system, and have the following functions, namely, averaging the demodulated outputs 1730 and 1740, which are positive and negative pulse signals, and respectively outputting LPF outputs 1750 and 1760 from which the components of communication data based on DoS attacks, etc., are removed. LPF outputs 1750 and 1760 are as follows: Figure 4 As shown, the waveform becomes a smooth waveform by averaging the unevenness of the pulse, and becomes a waveform suitable for phase determination described below.
[0051] Phase determination circuit 177 has a function of determining the phase of arrival timing of regular communication data, and outputs a phase matching signal 18 corresponding to the determined phase. The determination method uses the positive and negative signs of LPF outputs 1750 and 1760. Figure 5 1 is a diagram showing the relationship between the output of each LPF 175 and 176 and the phase quadrant determination result in the communication phase determination circuit 17 according to the first embodiment. Figure 6 17 is a diagram showing the relationship between the demodulated signals 1710 and 1720 and the phase matching signal 18 corresponding to the quadrants 1, 2, 3, and 4 according to the first embodiment.
[0052] Specifically, if Figure 5 As shown, the quadrant (phase range) is determined by using the LPF outputs 1750 and 1760 to determine whether they are positive or negative. Figure 6 The illustrated correspondence selects the phase matching signal corresponding to the quadrant and outputs it as the phase matching signal 18 .
[0053] For details, refer to Figure 5 , when both LPF outputs 1750 and 1760 are positive, it is quadrant 1; when both LPF output 1750 and LPF output 1760 are positive, it is quadrant 2; when both LPF outputs 1750 and 1760 are negative, it is quadrant 3; and when both LPF output 1750 and LPF output 1760 are negative, it is quadrant 4. Figure 4 In the case where the LPF output 1750 is positive and the LPF output 1760 is negative in all periods, it is determined to be quadrant 4.
[0054] Then, refer to Figure 6, the phase matching signal corresponding to quadrant 4 is a pulse that becomes Hi at the timing when the demodulation signal 1710 rises and becomes Lo at the timing when the demodulation signal 1720 rises. Figure 4 The timing shown outputs a phase matched signal 18 .
[0055] According to the above configuration, the communication phase determination circuit 17 can receive the input of the characteristic value 14 and output the phase matching signal 18 in such a way that the communication data in the vicinity of the reception timing of the regular communication data contained therein passes. Even if the DoS attack has periodicity, as long as the frequency and phase do not match the regular communication data, it is determined as improper communication data and discarded.
[0056] Furthermore, the communication volume per unit time of the communication data input to the input terminal Tin via the network 3 is compared with the communication volume per unit time of the communication data output from the output terminal Tout to the protected object device 2. If the latter communication volume is small, it can be regarded as communication data of the DoS attack reduced by the communication relay device 1. Furthermore, the periodicity of the communication data input to the input terminal Tin via the network 3 and the periodicity of the communication data output from the output terminal Tout to the protected object device 2 are compared. If the latter periodicity is closer to the control processing cycle of the control device 4 and the protected object device 2, it can be regarded as communication data of the DoS attack reduced by the communication relay device 1. By displaying the relevant information on the display device, it is possible to identify a DoS attack using a copy of regular communication data and to perform defense against the DoS attack.
[0057] (Effects of the First Embodiment)
[0058] In the prior art, communication data is identified by feature quantities and deleted based on the data frequency of each feature quantity. Therefore, there is a disadvantage that not only the improper communication data such as DoS attacks are discarded, but also the regular communication data that should have been passed is discarded. In contrast, according to the present embodiment, communication data is identified by feature quantities and deleted based on the data frequency of each feature quantity. On this basis, the periodicity of regular communication data is focused on, and communication data matching the periodicity is not discarded. Therefore, notifications based on DoS attacks are reduced and system failures of the protected device are prevented, and the disadvantageous situation of discarding communication data that should have been relayed is avoided, and the resistance to various DoS attacks can be improved.
[0059] (Second Embodiment)
[0060] Next, a communication relay device 1B according to the second embodiment will be described. The communication relay device 1B according to the second embodiment can improve the resolution of the phase to be determined compared to the communication relay device 1 according to the first embodiment.
[0061] The communication relay device 1B of the second embodiment is provided with a communication phase determination circuit 17B instead of the communication phase determination circuit 17, as compared with the communication relay device 1 of the first embodiment. In the communication relay device 1B, the communication phase determination circuit 17B is the same as the communication relay device 1 except for the expanded configuration compared with the communication phase determination circuit 17. Therefore, the differences in the configuration of the communication phase determination circuit 17B will be described here.
[0062] Figure 7 1 is a block diagram showing the configuration of a communication phase determination circuit 17B according to the second embodiment. The difference from the communication phase determination circuit 17 according to the first embodiment is that demodulation signal generating circuits 1711 and 1721, demodulation circuits 1731 and 1741, and low-pass filters 1751 and 1761 are added, and the outputs of each are input to a phase determination circuit 177.
[0063] The four demodulated signals 1710, 1720, 1712, 1722 including the additional configuration are as follows: Figure 8 as well as Fig. 9 The signal processing contents in the demodulation circuits 1731, 1741 and the low-pass filters 1751, 1761 are the same as the signal processing contents in the demodulation circuits 173, 174 and the low-pass filters 175, 176.
[0064] Figure 8 17 is a diagram showing the relationship between the output of each of the LPFs 175 , 176 , 1751 , and 1761 and the phase quadrant determination result in the communication phase determination circuit 17B according to the second embodiment. Fig. 9 17 is a diagram showing the relationship between the demodulated signals 1710 , 1720 , 1712 , and 1722 and the phase matching signal 18 corresponding to the quadrants 1 , 2 , 3 , 4 , 5 , 6 , 7 , and 8 according to the second embodiment.
[0065] According to the above configuration, a total of four LPF outputs 1750, 1760, 1752, and 1762 are obtained. The phase determination circuit 177 determines the phase according to the combination of the positive and negative signs of these LPF outputs and outputs the phase matching signal 18. As a specific method, Figure 8 As shown in FIG. 1 , the corresponding quadrant (phase range) among the eight quadrants is determined from the positive and negative relations of the four LPF outputs. Fig. 9 The illustrated correspondence selects the phase matching signal corresponding to the quadrant and outputs it as the phase matching signal 18 .
[0066] Reference Figure 8For example, when LPF outputs 1750, 1760, and 1752 are positive and LPF output 1762 is negative, it is quadrant 1. In this way, the quadrant is determined in accordance with the combination of the positive and negative signs of LPF outputs 1750, 1760, 1752, and 1762. Therefore, during the period when LPF outputs 1750, 1760, and 1752 are positive and LPF output 1762 is negative, it is determined to be quadrant 1.
[0067] Then, refer to Fig. 9 The phase matching signal corresponding to quadrant 1 is a pulse that becomes Hi at the timing when the demodulation signal 1712 rises and becomes Lo at the timing when the demodulation signal 1722 rises. The phase matching signal 18 is output at the timing of this pulse.
[0068] (Effects of the Second Embodiment)
[0069] According to this configuration, the resolution of the phase to be determined can be increased by two times compared to the communication relay device 1 of the first embodiment, and the amount of DoS attack communication that is mixed with regular communication and passes can be reduced.
[0070] (Variation of the embodiment)
[0071] In addition, the number of demodulation signals used in the first embodiment is two, and in the second embodiment is four, but the present invention is not limited to these numbers, and any number of demodulation signals can be used. The resolution of the phase determined by using more demodulation signals can be further improved. The number of demodulation signals is set to 2 in consideration of orthogonality. n (n is a natural number), it is more preferable to stagger the phases of the demodulated signals by 180 / 2 n The number of demodulated signals can be selected to be the optimal number by taking into account the trade-off between the circuit scale and the resolution corresponding to the number of demodulated signals.
[0072] In addition, in the first and second embodiments described above, the demodulation signal generated by the demodulation signal generating circuit is described as an example of a predetermined period, but the invention is not limited thereto. The period of regular communication data can be learned from the reception timing of the communication data, and the learned period can be set as the period of the demodulation signal, that is, the predetermined period representing the periodicity of regular communication can be learned based on the reception timing of the communication data. In this way, even when the control period changes, the standard communication period for judging the abandonment of the communication data can be automatically followed and automatically changed.
[0073] The present invention is not limited to the above-mentioned embodiments, and includes various modified examples. For example, the above-mentioned embodiments are the contents described in detail in order to explain the present invention in an easy-to-understand manner, and are not necessarily limited to all the structures described. In addition, as long as there is no contradiction, it is also possible to replace a part of the structure of a certain embodiment with the structure of other embodiments, and to add the structure of other embodiments to the structure of a certain embodiment. In addition, for a part of the structure of each embodiment, the structure can be added, deleted, replaced, integrated, or dispersed. In addition, the structure and processing represented by the embodiment can be appropriately dispersed, integrated, or interchanged based on processing efficiency or installation efficiency.
[0074] Description of Reference Numerals
[0075] 1, 1B: communication relay device, 2: protected device, 3: network, 4: control device, 9: attack device, 13: feature value extraction circuit, 15: average frequency determination circuit, 17, 17B: communication phase determination circuit, 19: discarding circuit.
Claims
1. A communication relay device that relays communication data received via a communication network to a protected device, wherein: have: a feature value extraction unit that extracts feature values from the communication data; a frequency determination unit that calculates an average value of the frequency of each feature value extracted by the feature value extraction unit, and outputs a cutoff signal when the average value of the frequency of each feature value is greater than a threshold value of each feature value; a phase determination unit that determines, for each feature value extracted by the feature value extraction unit, whether the reception timing of the communication data matches a range of phases corresponding to a predetermined period, and outputs a phase matching signal if the timing matches; as well as A discarding unit discards the communication data when the disconnection signal is output from the frequency determination unit and the phase matching signal is not output from the phase determination unit.
2. The communication relay device according to claim 1, characterized in that: The discarding unit discards the communication data when the phase matching signal is not output when the disconnection signal is output, and relays the communication data to the protected device without discarding the communication data when the phase matching signal is output.
3. The communication relay device according to claim 1, characterized in that: The prescribed period is learned based on the reception timing of the communication data.
4. The communication relay device according to claim 1, characterized in that: The phase range corresponding to the predetermined period is detected using synchronous demodulation that inverts the sign of the value of the demodulation target signal according to the sign of the synchronous demodulation signal.
5. The communication relay device according to claim 4, characterized in that: In the synchronous demodulation, the 180 / 2 n Phase shift of 2 n The synchronous demodulation signal is n, wherein n is a natural number.
6. The communication relay device according to claim 5, characterized in that: n=1。 7. The communication relay device according to claim 5, characterized in that: n=2。 8. The communication relay device according to claim 1, characterized in that: The frequency determination unit calculates the magnitude of the output of a low-pass filter of a first-order lag system that uses the reception timing of the communication data of each of the characteristic values as a pulse input as an average value of the frequency of each of the characteristic values.
9. The communication relay device according to claim 1, characterized in that: The magnitude of the output when the reception timing of the communication data for each of the characteristic values received within a certain period of time is used as a pulse input to a low-pass filter of a first-order lag system is calculated as a threshold value for each of the characteristic values.
10. A communication relay method, which is performed by a communication relay device that relays communication data received via a communication network to a protected device, wherein: The following processing is included: extracting a feature value from the communication data, The average value of the frequency of each of the extracted feature values is calculated, and a cutoff signal is output when the average value of the frequency of each feature value is greater than a threshold value of each feature value. For each of the extracted feature values, determining whether the reception timing of the communication data matches the range of the phase corresponding to the prescribed period, and outputting a phase matching signal if it matches, When the disconnection signal is output and the phase matching signal is not output, the communication data is discarded.
Citation Information
Patent Citations
Network system preventing distributed denial-of-service attack
JP2004318742A
Phase error detection device, method for detecting phase error, and receiving device
CN104205762A
Traffic classification system and method, and program, and abnormal traffic detection system and method
JP2010283668A
