Interference detection device, interference detection method, and program

Abstract

To provide an interference detection device, an interference detection method, and a program that allow users to determine the priority of interference to be dealt with in communication.SOLUTION: An interference detection device 10 according to an embodiment includes: a signal state model learning unit 11 that learns a signal state model using a learning signal in communication signals; an acquisition unit 12 that acquires a detection target signal in the communication signals; an interference detection unit 13 that detects interference in the detection target signal; and a severity calculation unit 14 that calculates the severity of interference in the detection target signal using the detection target signal, an interference detection result of the interference detection unit 13, and the learned signal state model.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to an interference detection device, an interference detection method, and a program. [Background technology]

[0002] In communication, techniques have been developed to discover and address degradation of communication quality due to interference, etc. For example, Patent Document 1 describes a communication quality degradation prediction system that predicts the amount of future degradation of communication quality inside and outside a communication section or the presence or absence of occurrence of degradation of communication quality, using a learning model related to degradation of communication quality in a certain communication section and degradation of communication quality outside the communication section. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] International Publication No. 2019 / 116417 Summary of the Invention [Problem to be solved by the invention]

[0004] Some types of interference in communication affect signals, while others have little effect. If a detection device detects both types of interference in communication, the user must deal with both types of interference, which can cause problems such as delays in dealing with the interference that affects signals and an increase in the time required to deal with the interference.

[0005] In view of the above-mentioned problems, an object of the present disclosure is to provide an interference detection device, an interference detection method, and a program that enable a user to determine the priority of interference that should be addressed in communication. [Means for solving the problem]

[0006] An interference detection device of one aspect of the present embodiment includes a signal state model learning unit that learns a signal state model using a learning signal in a communication signal, an acquisition unit that acquires a detection target signal in the communication signal, an interference detection unit that detects interference in the detection target signal, and a severity calculation unit that calculates the severity of interference in the detection target signal using the detection target signal, the interference detection result of the interference detection unit, and the learned signal state model.

[0007] An interference detection method according to one aspect of the present embodiment is one in which an interference detection device executes the following steps: a signal state model learning step of learning a signal state model using a learning signal in a communication signal; an acquisition step of acquiring a detection target signal in the communication signal; an interference detection step of detecting interference in the detection target signal; and a severity calculation step of calculating the severity of interference in the detection target signal using the detection target signal, the interference detection result, and the learned signal state model.

[0008] A program of one aspect of this embodiment causes a computer to execute, as an interference detection method, a signal state model learning step of learning a signal state model using a learning signal in a communication signal, an acquisition step of acquiring a detection target signal in the communication signal, an interference detection step of detecting interference in the detection target signal, and a severity calculation step of calculating the severity of interference in the detection target signal using the detection target signal, the interference detection result, and the learned signal state model. Effect of the Invention

[0009] According to the present disclosure, it is possible to provide an interference detection device, an interference detection method, and a program that enable a user to determine the priority of interference that should be addressed in communication. [Brief description of the drawings]

[0010] [Figure 1] 1 is a block diagram showing an example of an interference detection device according to a first embodiment. [Diagram 2] 5 is a flowchart showing an example of a process performed by the interference detection device according to the first embodiment. [Diagram 3] FIG. 11 is a block diagram showing an example of an interference detection system according to a second embodiment. [Figure 4] FIG. 11 is a state transition diagram showing an example of a signal state model according to the second embodiment. [Diagram 5] 13 is a graph showing an example of the severity of interference according to the second embodiment; [Figure 6A] 13 is a flowchart showing an example of a process for learning a signal state model in the interference detection device according to the second embodiment. [Figure 6B] 13 is a flowchart showing an example of a process for determining the presence or absence of interference according to the second embodiment. [Figure 6C] 13 is a flowchart showing an example of a process for calculating a severity level according to the second embodiment; [Figure 6D] 13 is a flowchart showing an example of a process for calculating a severity level according to the second embodiment; [Figure 7] FIG. 2 is a block diagram showing an example of a hardware configuration of an apparatus according to each embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0011] First embodiment Hereinafter, a first embodiment of the present invention will be described with reference to the drawings. Fig. 1 is a block diagram showing an example of an interference detection device. The interference detection device 10 is provided on the receiving side of wireless or wired communication, and detects interference in a signal related to the communication (communication signal). The interference detection device 10 in Fig. 1 includes a signal state model learning unit 11, an acquisition unit 12, an interference detection unit 13, and a severity calculation unit 14. Each unit (each means) of the interference detection device 10 is controlled by a control unit (controller) not shown. Each unit will be described below.

[0012] The signal state model learning unit 11 learns the signal state model by using a learning signal in the communication signal. The communication signal is a signal that can take a plurality of signal states, and for example, in the case of a digital signal, it has a binary state of 1 (on state) or 0 (off state), but it may be a signal having three or more states. The learning signal is a signal used for learning to determine the signal state of a detection target signal, which will be described later, among the communication signals. The learning signal may be one acquired by the interference detection device 10 from an acquisition unit 12, which will be described later, or may be one stored in advance by the interference detection device 10.

[0013] The signal state model is a state transition model in which, for multiple states that a learning signal can take, when the signal is in a certain state in a certain time slot (a time period that is one unit of the signal), the transition probability that the signal will transition to another state different from the state in the adjacent time slot is defined. For example, when a signal can take two states, a first state and a second state, the signal state model defines the transition probability of transitioning from the first state to the second state, the transition probability of transitioning from the second state to the first state, and the transition probability of each state being continued. In addition, when a signal can take three states, A, B, and C, the signal state model defines the transition probability of transitioning from A to B, the transition probability of transitioning from B to A, the transition probability of transitioning from A to C, the transition probability of transitioning from C to A, the transition probability of transitioning from B to C, the transition probability of transitioning from C to B, and the transition probability of each state being continued.

[0014] The acquisition unit 12 is an interface that acquires a detection target signal in a communication signal, and outputs the acquired detection target signal to the interference detection unit 13 and the severity calculation unit 14. The multiple signal states that the detection target signal can take are common to the multiple signal states that the learning signal can take. The detection target signal may be a signal in the same or adjacent frequency band as the learning signal. For example, the detection target signal and the learning signal may be wireless signals received by the same device or adjacent devices, and the detection target signal may be received by the interference detection device 10 in a different period from the learning signal. Alternatively, a part of the transmission path from the transmission source of the detection target signal to the interference detection device 10 may be a wired or wireless path that is common to or adjacent to a part of the transmission path from the transmission source of the learning signal to the interference detection device 10. For example, the learning signal and the detection target signal may be transmitted to the interference detection device 10 via a common LAN (Local Area Network) cable connected to the interference detection device 10. Here, "proximity" indicates a condition in which the same interference occurs in the detection target signal and the learning signal. For example, when the detection target signal is a signal in a frequency band close to that of the learning signal, it means that the learning signal can be received in a bandwidth in which the detection target signal can be received.

[0015] The interference detection unit 13 detects interference in a detection target signal. The period during which signal interference is detected is, for example, a period that does not overlap with a learning period, and the interference detection unit 13 can detect in which time slot of the detection target signal the interference exists.

[0016] The severity calculation unit 14 calculates the severity of interference in the detection target signal using the detection target signal, the interference detection result of the interference detection unit 13, and the learned signal state model. In detail, the severity calculation unit 14 applies the learned signal state model to the data of the signal state in at least any period (one or more time slots) before or after the time series in one or more time slots in which interference occurs in the detection target signal. Thereby, the severity calculation unit 14 determines the severity to be higher as the probability that the signal state in the time slot in which interference occurs is on or the period in which interference occurs is longer. When the probability that the signal state in the time slot in which interference occurs is on is high, it is highly likely that the interference overlaps with the on state of the signal, and therefore it is considered that the interference is highly likely to affect the signal. Also, when the period in which interference occurs is long, it is highly likely that the interference overlaps with the on state of the signal, or the period in which the interference overlaps with the on state of the signal is long. In this case, it is also considered that the interference is highly likely to affect the signal.

[0017] 2 is a flowchart showing an example of a representative process of the interference detection device 10, and this flowchart will be used to explain the process of the interference detection device 10. First, the acquisition unit 12 of the interference detection device 10 learns a signal state model using a learning signal (step S11; signal state model learning step).

[0018] Next, the acquisition unit 12 acquires the detection target signal (step S12; acquisition step). The interference detection unit 13 detects interference in the detection target signal (step S13; interference detection step). Then, the severity calculation unit 14 calculates the severity of the interference in the detection target signal using the detection target signal, the interference detection result of the interference detection unit 13, and the learned signal state model (step S14; severity calculation step). The interference detection device 10 can notify the user of the calculated severity by a display unit, a voice output unit, etc.

[0019] In this manner, the interference detection device 10 can calculate the severity of interference in a detection target signal using the signal state model. Therefore, a user can immediately understand whether the interference is serious or not based on the calculated severity, and can determine the priority of the interference to be dealt with in communication.

[0020] Embodiment 2 A second embodiment of the present invention will be described below with reference to the drawings. In the second embodiment, a specific example of the interference detection device 10 described in the first embodiment will be disclosed.

[0021] 3 is a block diagram showing an example of an interference detection system 20. The interference detection system 20 is a device constituting a wireless signal interference detection system, and includes a receiver 21, a signal state model learning unit 22, a signal state calculation unit 23, an interference detection unit 24, a severity calculation unit 25, and a severity presentation unit 26. Each unit will be described in detail below.

[0022] The receiving unit 21 is an interface that receives a wireless signal R (a communication signal in the first embodiment) from a transmitting device (not shown). The wireless signal R is transmitted in a predetermined frequency band and is a signal having two values, an on state or an off state. The receiving unit 21 outputs signal information of the wireless signal R (learning signal) during the acquired learning period to the signal state model learning unit 22, and outputs signal information of the wireless signal R (a detection target signal) during the acquired detection period to the signal state calculation unit 23 and the interference detection unit 24. Note that the learning period is a period that precedes the detection period in the time series. In this way, the learning signal and the detection target signal each contain signal information of a different wireless signal R. In the learning signal, the on state and off state of the signal are analyzed by the signal state model learning unit 22 and become known. In contrast, in the detection target signal, the on state and off state of the signal are derived by calculation by the signal state calculation unit 23.

[0023] The signal state model learning unit 22 uses the acquired learning signal to learn a signal state model based on a predefined signal state model. Here, the signal state model is a model used to calculate whether the value of the signal to be calculated in each time slot is in an on state or an off state. The signal state model has signal states, transitions, and transition probabilities determined by learning, which are predefined by the signal state model learning unit 22. The signal state model learning unit 22 learns the transition probabilities using information on the signal states in the learning signal.

[0024] The signal to be calculated is a signal that is interfered with in the interference detection system, such as voice in the case of voice communication, data in the case of data communication, and management information in the case of management communication. The signal state model learning unit 22 can determine whether or not a signal in an on state (on signal) is included by demodulating the received signal or detecting that the signal is received in a certain frequency band with a strength equal to or higher than a certain reception level, but the signal may be determined by other methods. In this way, the signal state model learning unit 22 learns the signal state model before the detection target signal is transmitted. The learned signal state model is output to the signal state calculation unit 23. The learned signal state model may also be stored in a storage unit (not shown).

[0025] FIG. 4 is a state transition diagram showing an example of a signal state model generated by the signal state model learning unit 22. In the signal state model in FIG. 4, the on state (S1) and off state (S2) of the signal, the transition probability for each state to continue, and the transition probability for each state to transition to a different state are set. In FIG. 4, when the signal is on in the time slot to be analyzed, the probability that the signal will be on at the next time slot in the time series (probability that the state will continue) is set as p(S1), and the probability that the signal will be off (probability that the state will transition) is set as 1-p(S1). On the other hand, when the signal is off in a certain time slot, the probability that the signal will be off at the next time slot in the time series is set as p(S2), and the probability that the signal will be off is set as 1-p(S2). The signal state model learning unit 22 determines these four transition probabilities by learning.

[0026] The signal state model learning unit 22 may define only the state of the transition source as a parameter for determining each transition probability, or may further define other parameters. The other parameters are, for example, the duration (number of time slots) of the current signal state and the number of transitions. The signal state model learning unit 22 determines each transition probability by setting such parameters through learning. Also, the states of the signal state model, the number of possible states, and the transition method are not limited to those shown in FIG. 4.

[0027] Returning to Fig. 3, the explanation will be continued. Signal state calculation unit 23 calculates the signal state of the detection target signal using the learned signal state model generated by signal state model learning unit 22 and the signal information of the detection target signal output from receiving unit 21. For example, in the signal state model shown in Fig. 4, the signal state is an on signal or an off signal, so that whether the signal state of a certain time slot is an on signal or an off signal in the signal information of a period without interference is derived by calculation using the signal state model. Then, the derived signal state is output to severity calculation unit 25.

[0028] The interference detection unit 24 performs interference determination using the received signal information. Any existing technology can be used for this interference determination, so a detailed description will be omitted. As an example of the existing technology, a frequency mask technology for interference determination or a technology for classification or identification by some kind of learning may be used. As a method for generating this frequency mask, a machine learning algorithm, for example, One-Class SVM (Support Vector Machine), which is an outlier detection method, can be used. One-Class SVM is a general method, so it will be omitted. The interference detection unit 24 performs interference determination over the entire time series of the signal, and outputs the result of determining whether or not there is interference in each period or time slot of the signal to the severity calculation unit 25.

[0029] The severity calculation unit 25 calculates the severity (significance) of interference in the detection target signal using the signal state calculated by the signal state calculation unit 23 and the interference detection result detected by the interference detection unit 24. Here, the severity is a factor considering the influence on wireless communication, meaning whether or not an ON signal overlaps with interference and the degree of the overlap. In other words, the more ON signals there are during the period when interference occurs, or the longer the interference occurs, the higher the severity of the interference. The method of calculating the severity will be further explained below.

[0030] Fig. 5 is a graph showing an example of the severity of interference. Fig. 5 shows on signals O1 to O4 and interferences I1 to I4. Interferences I1 and I2 are not serious interferences because they do not overlap with on signal O. On the other hand, interferences I3 and I4 are serious interferences because they overlap with on signals O3 and O4, respectively. Also, the duration of interference I4 in on signal O4 is longer than the duration of interference I3 in on signal O3, so interference I4 is more serious than interference I3.

[0031] Since the user of the wireless communication on the interfered side does not intentionally cause interference, it is generally not possible to know in advance the characteristics of the interfering signal. In other words, it is not possible to know the degree of overlap between the on signal and the interference shown in Fig. 5. For this reason, during a period in which interference occurs in the signal information, the severity calculation unit 25 needs to estimate the signal state using the learned signal state model.

[0032] (i) For example, the severity calculation unit 25 can calculate the probability of the ON state during the period during which it is determined that interference occurs, by using the transition probability defined in the learned signal state model and the signal state of the time slot immediately preceding the period during which it is determined that interference occurs. When it is determined that interference occurs during one time slot (i.e., the period for one state transition determination), the probability P of the ON state during the period during which interference occurs is as follows in the model shown in FIG. (1) If the time slot immediately before the period of interference was in the off state, P = 1-p(S2) (2) If the signal was on in the time slot immediately before the period of interference, P = p(S1) The severity calculation unit 25 calculates the severity based on this probability P. However, the severity calculation unit 25 can also calculate the probability P of going into the on state during the period determined to be interference by using the transition probability defined in the learned signal state model and the signal state of the time slot immediately following the period determined to be interference.

[0033] (ii) As another example, when the period determined to be interference has a time length spanning multiple time slots, the severity calculation unit 25 may estimate a state transition path during the period of interference using the learned signal state model and the signal state for a certain period before and after the period determined to be interference. In this case, the severity calculation unit 25 may calculate the number of time slots (time length) in the on state, or the number of consecutive states in which the signal is on, as the severity based on the state transition in the estimation result. In addition, the severity calculation unit 25 may calculate the severity by adding up the seriousness of multiple state transition paths with high transition probability. Calculations based on the above examples shown in (i) and (ii) will be shown as calculation examples described later.

[0034] 3, the description will continue. The seriousness presenting unit 26 is an interface that presents the seriousness calculated by the seriousness calculation unit 25 to the user of the interference detection system 20. For example, the seriousness presenting unit 26 is composed of a display unit and an audio output unit.

[0035] 6A to 6D are flowcharts showing an example of a representative process of the interference detection system 20, and the flowcharts will be used to explain a series of processes of the interference detection system 20. Note that details of the processes executed by each unit are as described above, and will therefore be omitted where appropriate.

[0036] First, the process of learning the signal state model of the interference detection system 20 shown in Fig. 6A will be described. The receiver 21 receives the wireless signal R during a learning period to obtain signal information of the learning signal (step S21). Next, the signal state model learning unit 22 learns the signal state model using the signal information obtained in step S21 (step S22). The signal state model learning unit 22 stores the learned signal state model in a storage unit (not shown) of the interference detection system 20 (step S23).

[0037] Next, the process of learning the signal state of the detection target signal of the interference detection system 20 shown in Fig. 6B will be described. The receiver 21 obtains signal information of the detection target signal by receiving the wireless signal R during the detection period (step S31). The signal state calculation unit 23 calculates the signal state for each time slot of the detection target signal using the signal information obtained in step S31 and the signal state model learned in step S22 (step S32). The signal state calculation unit 23 outputs the calculated signal state to the severity calculation unit 25. Note that the signal state calculation unit 23 may store the signal state in a storage unit (not shown) of the interference detection system 20.

[0038] Next, the process of determining whether or not there is interference by the interference detection system 20 shown in Fig. 6C will be described. The interference detection unit 24 determines whether or not there is interference for each time slot of the detection target signal (step S41). The interference detection unit 24 outputs the time information of the time slot and information indicating the presence or absence of interference in that time slot to the severity calculation unit 25 as an interference detection result.

[0039] If interference exists in the time slot to be detected (Yes in step S41), the severity calculation unit 25 calculates the severity of the interference during the period when the interference exists, using the signal state in that time slot calculated in step S32 and the interference detection result detected by the interference detection unit 24. The details of this will be described later. Then, the severity presentation unit 26 presents the severity (step S42).

[0040] Then, the interference detection unit 24 stores the interference determination result of step S41 in a storage unit (not shown) of the interference detection system 20 (step S43). Note that the processes of steps S42 and S43 may be performed in either order, or may be performed in parallel. Even if there is no interference in the time slot to be detected (No in step S41), the interference detection unit 24 executes the process of step S43 to store the interference determination result in the storage unit.

[0041] Finally, a detailed example of the process of step S42 shown in Fig. 6D will be described. First, the seriousness calculation unit 25 extracts the signal state immediately before the period in which it is determined that there is interference (step S51). Next, the seriousness calculation unit 25 determines the transition probability of the signal state extracted in step S51 to the ON state in the learned signal state model, and calculates the seriousness using the transition probability (step S52). The seriousness presentation unit 26 presents the seriousness calculated in step S52 to the user (step S53).

[0042] In this way, the interference detection system 20 can calculate and present the severity of interference in the detection target signal using the probability that the signal will be in the ON state during the period of interference based on the learned signal state model. Therefore, as in the first embodiment, the user can determine the priority of the interference that should be dealt with in communication.

[0043] Wireless communication using radio waves is used in various fields, and there are various technologies to determine whether interference occurs in such communication and to deal with the interference. Among such communications, particularly important communications such as police radio and railway radio are called important radio communications. If interference occurs in important radio communications, it could develop into a situation that puts human lives at risk, so it is very important to detect (monitor) anomalies in the emission status of the radio waves used for such communications.

[0044] As a kind of unsupervised learning for detecting interference with an unknown received signal, there is a method for detecting interference by learning a known signal that is not subjected to interference, and using factors such as the distance and similarity between the unknown signal and the learned known signal. Here, factors such as distance and similarity do not take into account the influence of interference on wireless communication actually used. Therefore, when an interference detection system automatically detects interference using such a method, the interference detection system detects a large amount of interference, including interference that does not affect wireless communication and interference whose influence is so small that it can be ignored. As a result, a problem occurs in that a user of the interference detection system does not know which interference should be given priority for cause analysis and solution among the large amount of interference detected.

[0045] In response to this, the interference detection system 20 can present to the user the severity reflecting the effect (the actual degree of quality degradation) of the detected interference on the on-state signal. This allows the user to determine which interference should be dealt with, such as by analyzing the cause and solving it, as a priority. This makes it possible to deal with interference efficiently in wireless communication.

[0046] [Calculation example] Next, the implementation of the interference detection system 20 will be described using a specific calculation example. In the following example, a case will be described in which the interference detection system 20 receives radio waves in the 400 MHz frequency band as the wireless signal R that is affected by interference. Here, the 400 MHz frequency band means that a frequency width of 1 MHz, centered around 400 MHz, is used for radio wave transmission.

[0047] First, the user sets the frequency band of the receiver 21 of the interference detection system 20 as described above. After that, the receiver 21 receives radio waves of 400 MHz ± 0.5 MHz as a wireless signal R for about 100 seconds, which is a learning period. The signal state model learning unit 22 obtains time series data of an IQ signal for this received learning signal. Here, when communication is in an on state, the reception level is high, and when communication is in an off state, the reception level is low. Therefore, the signal state model learning unit 22 converts the time series data of the amplitude in the obtained time series data into state time series data in which the absolute value of the amplitude exceeds a certain threshold value and the state is off when the absolute value does not exceed the threshold value. The threshold value for determining the on state and the off state is stored in the storage unit of the interference detection system 20, and the signal state model learning unit 22 uses it during learning. In this example, it is assumed that 101 samples (1 sample: 1 time slot to be analyzed) of state time series data have been obtained.

[0048] [Calculation example 1] First, an example in which the signal state model learning unit 22 calculates the signal state model shown in Fig. 4 will be shown. Here, it is assumed that in the time series data of the learning signal, there are 20 transitions starting in the ON state, and 18 transitions in which the ON state continues. In this case, p(S1) = 18 / 20 = 0.9 (1) Here, assuming that the number of transitions starting in the OFF state is 80, and the number of transitions in which the OFF state continues is 78, p(S2) = 78 / 80 = 0.975 (2) By calculating p(S1) and p(S2) in this manner, signal state model learning unit 22 learns the signal state model.

[0049] Next, the receiver 21 of the interference detection system 20 receives radio waves with the same frequency band as the learning signal set, thereby acquiring time series data of the IQ signal, which is the detection target signal. The signal state calculation unit 23 converts this time series data into state time series data in which the on and off states are set, similar to the learning signal.

[0050] Furthermore, the interference detection unit 24 performs interference determination for the detection target signal using the time series data of the IQ signal acquired by the receiving unit 21. Here, it is assumed that the interference detection unit 24 has determined that there is interference in the time zone of the 50th sample in the state time series data of the detection target signal.

[0051] The severity calculation unit 25 calculates the severity of the interference of the 50th sample as follows: If the state of the 49th sample of the state time-series data of the detection target signal is the on state (S1), referring to the learned signal state model, the probability p(S1) that S1 will continue is 0.9 as shown in (1). Therefore, the severity calculation unit 25 calculates the severity of this interference as 0.9.

[0052] [Calculation example 2] In this calculation example 2, it is assumed that the interference detection unit 24 determines that interference exists in the consecutive time periods of the 50th and 51st samples of the state time-series data of the detection target signal. Here, the learned signal state model is as shown in calculation example 1. Since the probability p(S1) that the signal is in the on state (S1) is 0.9, the severity calculation unit 25 calculates the severity of the interference for the 50th sample to be 0.9 in the same manner as calculation example 1.

[0053] Furthermore, the interference detection unit 24 determines that the detection target signal is in the ON state (S1) at the 50th sample and the ON state of the signal continues at the 51st sample as follows: p1 = 0.9 × 0.9 (3) In addition, the probability p2 that the detection target signal is in the OFF state (S2) at the 50th sample and the signal transitions to the ON state at the 51st sample is p2 = 0.1 × 0.025 (4) From the above, the interference detection unit 24 calculates the severity of the interference in the 51st sample as 0.8125 by taking the sum of (3) and (4).

[0054] [Calculation example 3] In this calculation example 3, the severity calculation unit 25 calculates p1 and p2 as shown in (3) and (4) by performing calculations similar to those in calculation example 2. Then, the severity calculation unit 25 calculates the overall severity of interference in the 50th sample and the 51st sample (i.e., the severity in the detection target signal) as the sum of the severity in both samples, 0.9 + 0.8125 = 1.7125.

[0055] However, the severity calculation unit 25 may calculate the average value of the severity in both samples as the severity in the detection target signal. Furthermore, the severity presentation unit 26 may present the duration of the interference determined by the interference detection unit 24 as "interference occurs" together with the severity calculated by the severity calculation unit 25. The duration may be defined as the number of samples, for example. In this case, since it is determined that interference occurs in the 50th sample and the 51st sample, the duration is 2.

[0056] [Calculation example 4] In this calculation example 4, the time length L during which the signal state of the target sample continued in the past is further defined as a parameter for determining the transition probability of the signal state model. In this example, the time length L is the number of samples that continue in a certain state on the state time-series data of the detection target signal. For example, if the signal state of the target sample is an on state and the signal state immediately before that was an off state, L is 1. On the other hand, if the signal state of the target sample is an off state, the signal state one state before that was an off state, and the signal state two states before that was an on state, L is 2.

[0057] Also, when an on state (S1) continues for a time length L, the probability that the on state will continue in the next sample is assumed to be p(S1, L), and when an off state (S2) continues for a time length L, the probability that the off state will continue in the next sample is assumed to be p(S2, L). Signal state model learning unit 22 learns a signal state model with this time length L as a parameter. In this learned signal state model, p(S1,1) and p(S1,4) are assumed to be as follows. p(S1,1)=1.0 (5) p(S1,4) = 0.0 (6)

[0058] Here, as in Calculation Example 1, it is assumed that interference is detected in the time period of the 50th sample of the state time-series data of the detection target signal, the state of the 49th sample is the on state (S1), and L=1 at the 49th sample. By referring to the learned signal state model, the severity calculation unit 25 determines that the probability p(S1,1) that the on state continues is 1.0, as shown in (5). As a result, the severity of interference at the 50th sample is calculated to be 1.0.

[0059] Furthermore, assume that interference is detected in the time period of the 65th sample of the state time-series data of the detection target signal, the state of the 64th sample is the on state (S1), and L=4 at the 64th sample. The severity calculation unit 25 determines that the probability p(S1,4) that the on state continues is 0.0, as shown in (6), by referring to the learned signal state model. As a result, the severity of interference at the 64th sample is calculated to be 0.0.

[0060] As described above, if it is considered that the on state will continue until the period during which it is determined that interference occurs, the severity will be 1, which is the maximum value. In this case, the user can recognize that the interference should be dealt with as a priority. On the other hand, if it is not considered that the on state will continue until the period during which it is determined that interference occurs, the severity will be 0. In this case, the user can recognize that the priority for dealing with the interference is low, or that there is no need to deal with it.

[0061] [Calculation example 5] In this calculation example 5, similarly to calculation example 4, the signal state model learning unit 22 defines the time length L as a parameter and learns a communication state model. Then, similarly to calculation example 2, it is assumed that interference is detected in the time period between the 50th and 51st samples of the state time-series data of the detection target signal, and the signal state of the 49th sample is the on state (S1). Furthermore, it is assumed here that the signal state of the 52nd sample is the on state (S1).

[0062] At this time, the severity calculation unit 25 determines that four types of state transition paths are possible for the 50th and 51st samples by referring to the communication state model. The four types of paths are path 1 [S1, S1, S1, S1], path 2 [S1, S1, S2, S1], path 3 [S1, S2, S1, S1], and path 4 [S1, S2, S2, S1]. Note that, for each path, [R1, R2, R3, R4], R1 is the signal state of the 49th sample, R2 is the signal state of the 50th sample, R3 is the signal state of the 51st sample, and R4 is the signal state of the 54th sample.

[0063] In this calculation example 5, the transition probabilities p(S1,1), p(S1,2), p(S1,3), p(S2,1), and p(S2,2) related to the calculation defined in the communication state model are respectively p(S1,1)=0.9 (7) p(S1,2)=0.9 (8) p(S1,3)=0.7 (9) p(S2,1)=1.0 (10) p(S2,2)=0.9 (11) Here, the seriousness calculation unit 25 calculates the transition probability in each of the paths 1 to 4 using (7) to (11) as follows: (Route 1) 0.9 × 0.9 × 0.7 = 0.567 (12) (Route 2) 0.9 × 0.1 × 0.0 = 0.0 (13) (Route 3) 0.1 × 0.0 × 0.9 = 0.0 (14) (Route 4) 0.1 × 1.0 × 0.1 = 0.01 (15) Based on the calculations of (12) to (15), the seriousness calculation unit 25 determines that path 1 has the highest transition probability. As a result, the seriousness calculation unit 25 estimates that the transition path of the signal state during the period between the 50th sample and the 51st sample when interference was detected was path 1. By determining that two samples of the on state (S1) were included during the period when interference was present on path 1, the seriousness calculation unit 25 calculates the overall seriousness of this period as 2, which indicates that the period in which the on state was included was two samples. Note that the seriousness calculation unit 25 may determine the period in which the on state was included for the entire period of path 1, which is the path with the highest transition probability, and calculate the seriousness of that period as "4".

[0064] In the above example, the severity calculation unit 25 searches for paths by exhaustive search, but other methods may be used for path search. Also, the severity calculation unit 25 may use the length of time that the path was in the on state as the severity instead of the number of samples that were in the on state during the entire period or the period with interference in the path with the highest transition probability. For example, in the above example, if the length of time of one sample is 1 second, the severity is 2.

[0065] As another example, the severity calculation unit 25 may regard a continuous ON state for one or more samples as a unique signal in each path, determine the number of unique signals (1 for paths 1 to 3, 0 for path 4) in the period determined to have interference, and calculate the number of unique signals in the period determined to have interference for the path with the highest transition probability as the severity. Alternatively, the severity calculation unit 25 may determine the number of unique signals for the path with the highest transition probability not in the period determined to have interference, but in the entire period of the path including before and after the period determined to have interference. For example, as described above, it is assumed that interference is detected in the time zone of the 50th and 51st samples of the state time series data of the detection target signal, and the signal state of the 49th and 52nd samples is the ON state (S1). Then, when the transition probability P related to the calculation is set as (7) to (11), it is determined that the path 1 of (12) has the highest transition probability among the paths including before and after the period determined to have interference. Therefore, the seriousness calculation unit 25 determines that the number of unique signals in the path 1 is "1", and calculates the seriousness as "1".

[0066] The severity presenting unit 26 presents to the user the severity of interference in each sample calculated as in the calculation examples 1 to 5, or the severity in the detection target signal taking all interference into consideration. The severity presenting unit 26 may present the period during which the interference detected by the interference detection unit 24 exists in association with the severity. Furthermore, as in the calculation examples 2, 3, and 5, when it is determined that interference exists over a plurality of samples, the severity presenting unit 26 may present the length of the period during which the interference exists (e.g., the sample length) together with the severity. Furthermore, when it is determined that interference exists over a plurality of samples, the severity presenting unit 26 may present the severity for each sample.

[0067] As described above, the severity calculation unit 25 can calculate the severity of interference for each of a plurality of samples based on the state transition path of the signal state in a plurality of samples (time slots) in the learned signal state model. This allows the user to determine and deal with the interference that should be prioritized from among the interferences occurring in a plurality of samples, thereby enabling efficient handling.

[0068] Here, the severity calculation unit 25 can also calculate the severity of interference in the detection target signal based on the calculated severity of each of the multiple samples. Therefore, the user can determine whether or not to deal with the interference occurring in the entire multiple samples as a priority. For example, when the severity is high, the user can determine that it is necessary to deal with the interference as a priority, whereas when the severity is low, the user can determine that it is less necessary to deal with the interference.

[0069] In addition, in the signal state model, the signal state can be either an on state or an off state (i.e., the number of transition states of the signal state is 2). Therefore, the interference detection system 20 can calculate the severity of interference for a normal digital signal.

[0070] Here, the signal state model learning unit 22 can set the length of samples (time slots) in which the on state continues as a parameter of the state transition probability of the signal state model. The severity calculation unit 25 calculates the severity of interference in the detection target signal by using the parameter and the number of samples in which the on state continues immediately before the sample in which the interference detection unit 24 detects interference in the detection target signal. This allows the interference detection system 20 to present a more accurate severity.

[0071] Furthermore, the severity calculation unit 25 uses the signal state of the detection target signal immediately before and after the period in which the interference detection unit 24 detects interference and the learned signal state model to identify a state transition path having the maximum transition probability among multiple state transition paths of the signal state during that period. Then, at least one of the length of time the signal is in the ON state, the number of time slots in which the signal is in the ON state, the number of consecutive states in which the signal is in the ON state, or an expected value of any of these, during at least either the entire period of the identified state transition path or the period in which interference is detected, can be calculated as the severity of interference. This allows the interference detection system 20 to present the severity, which is an appropriate index according to the nature of the signal. Note that there may be multiple state transition paths having the maximum transition probability.

[0072] In addition, the learning signal and the detection target signal may be wireless signals in the same or adjacent frequency bands, which allows the interference detection system 20 to detect the severity of interference of a wireless signal in a certain frequency band without knowing the content of that signal in advance.

[0073] The present invention is not limited to the above-mentioned embodiment, and can be modified as appropriate without departing from the spirit of the present invention. For example, in the calculation examples 1 to 5 in the second embodiment, the seriousness presenting unit 26 calculates the state transition probability assuming that the state transition occurs in ascending order (chronological order) of the samples, but the state transition probability may be calculated assuming that the state transition occurs in descending order (opposite to the chronological order) of the samples. For example, in the calculation example 1, when the state of the 51st sample of the state time-series data of the detection target signal is the ON state (S1), the seriousness calculation unit 25 refers to the learned signal state model and specifies the probability that S1 will continue until the 50th sample. Then, the seriousness calculation unit 25 calculates the probability as the seriousness of the interference of the 50th sample.

[0074] In calculation example 5, the severity calculation unit 25 may use the calculated transition probability for each path to calculate the expected value of the number of samples that were in the ON state during the period in which interference was determined to exist (specifically, the time period between the 50th sample and the 51st sample) as the severity. For example, in the above calculation example 5, the expected value of the number of samples is (2*0.567+1*0.0+1*0.0+0*0.01)≒1.134 (16) Therefore, the severity is 1.134.

[0075] Alternatively, in calculation example 5, the severity calculation unit 25 may use the calculated transition probability for each path to calculate, as the severity, the expected value of the number of samples that were in the ON state not only during the period in which interference was determined to exist, but also during the entire period on the path including the periods before and after the period (specifically, the time period from the 49th sample to the 52nd sample). For example, in the above calculation example 5, the expected value of the number of samples is (4*0.567+3*0.0+3*0.0+2*0.01)≒2.27 (17) Therefore, the severity is 2.27. In this way, the severity calculation unit 25 may calculate the expected value of the number of samples for the period in which it is determined that interference occurs, or may calculate the entire period on the route including the period in which it is determined that interference occurs and the periods before and after the period.

[0076] 7 is a block diagram showing an example of a hardware configuration of an information processing device (signal processing device) for executing the processes of the above-described embodiments. Referring to FIG. 7, the information processing device 90 includes a signal processing circuit 91, a processor 92, and a memory 93.

[0077] The signal processing circuit 91 is a circuit for processing signals in accordance with the control of the processor 92. The signal processing circuit 91 may include a communication circuit for receiving a signal from a transmitting device.

[0078] The processor 92 performs the processing of the device described in the above embodiment by reading and executing software (computer programs) from the memory 93. As an example of the processor 92, one of a CPU (Central Processing Unit), an MPU (Micro Processing Unit), an FPGA (Field-Programmable Gate Array), a DSP (Demand-Side Platform), and an ASIC (Application Specific Integrated Circuit) may be used, or a plurality of them may be used in parallel.

[0079] The memory 93 is composed of a volatile memory, a non-volatile memory, or a combination thereof. The memory 93 is not limited to one, and a plurality of memories may be provided. The volatile memory may be, for example, a RAM (Random Access Memory) such as a DRAM (Dynamic Random Access Memory) or a SRAM (Static Random Access Memory). The non-volatile memory may be, for example, a ROM (Random Only Memory) such as a PROM (Programmable Random Only Memory) or an EPROM (Erasable Programmable Read Only Memory), a flash memory, or an SSD (Solid State Drive).

[0080] The memory 93 is used to store one or more instructions. Here, the one or more instructions are stored in the memory 93 as a group of software modules. The processor 92 can perform the processes described in the above-mentioned embodiments by reading and executing the group of software modules from the memory 93.

[0081] The memory 93 may include a memory built into the processor 92 in addition to a memory provided outside the processor 92. The memory 93 may also include a storage device disposed away from the processors constituting the processor 92. In this case, the processor 92 can access the memory 93 via an I / O (Input / Output) interface.

[0082] As described above, one or more processors included in the interference detection device 10 or the interference detection system 20 execute one or more programs including a set of instructions for causing a computer to execute the algorithm described with reference to the drawings. This processing realizes the signal processing method described in each embodiment.

[0083] The program includes a set of instructions (or software code) that, when loaded into a computer, causes the computer to perform one or more functions described in the embodiments. The program may be stored in a non-transitory computer-readable medium or a tangible storage medium. By way of example and not limitation, the computer-readable medium or tangible storage medium includes random-access memory (RAM), read-only memory (ROM), flash memory, solid-state drive (SSD) or other memory technology, CD-ROM, digital versatile disk (DVD), Blu-ray® disk or other optical disk storage, magnetic cassette, magnetic tape, magnetic disk storage or other magnetic storage device. The program may be transmitted on a transitory computer-readable medium or communication medium. By way of example and not limitation, the transitory computer-readable medium or communication medium includes electrical, optical, acoustic, or other forms of propagated signals.

[0084] Although the present disclosure has been described above with reference to the embodiments, the present disclosure is not limited to the above. Various modifications that can be understood by a person skilled in the art can be made to the configuration and details of the present disclosure within the scope of the disclosure. [Explanation of symbols]

[0085] 10 Interference detection device 11 Signal state model learning unit 12 Acquisition unit 13 Interference detection unit 14 Severity calculation unit 20 Interference Detection System 21 Receiving section 22 Signal state model learning section 23 Signal state calculation unit 24 Interference detection unit 25 Severity level calculation section 26 Severity level presentation section

Claims

1. a signal state model learning unit that learns a signal state model by using a learning signal in the communication signal; an acquisition unit that acquires a detection target signal in the communication signal; An interference detection unit that detects interference in the detection target signal; a severity calculation unit that calculates a severity of interference in the detection target signal by using the detection target signal, an interference detection result of the interference detection unit, and the learned signal state model. Interference detection device.

2. When the interference detection unit detects interference in a plurality of consecutive time slots of the detection target signal, the severity calculation unit calculates the severity of interference for each of the plurality of time slots based on a state transition path of a signal state in the plurality of time slots in the learned signal state model.

2. An interference detection device as claimed in claim 1.

3. The severity calculation unit calculates a severity of interference in the detection target signal based on the calculated severity of each of the plurality of time slots.

3. An interference detection device as claimed in claim 2.

4. The signal state model is such that the signal state can be either an on state or an off state. An interference detection device according to any one of claims 1 to 3.

5. the signal state model learning unit sets a time slot length during which the on state continues as a parameter of a state transition probability of the signal state model; the severity calculation unit calculates the severity of the interference in the detection target signal by using the parameter and a time slot length in which the on state continues in the detection target signal immediately before the time slot in which the interference detection unit detects interference; 5. An interference detection device as claimed in claim 4.

6. the severity calculation unit uses the signal state of the detection target signal immediately before and after the period in which the interference detection unit detects interference and the learned signal state model to identify a state transition path having a maximum transition probability among a plurality of state transition paths of the signal state during the period, and calculates, as the severity of the interference, at least one of the length of time in which the signal is in an on state, the number of time slots in which the signal is in an on state, the number of consecutive states in which the signal is in an on state, or an expected value of any of these, during at least either the entire period of the identified state transition path or the period in which the interference was detected; 5. An interference detection device as claimed in claim 4.

7. The learning signal and the detection target signal are radio signals in the same or adjacent frequency bands. An interference detection device according to any preceding claim.

8. a signal state model learning step of learning a signal state model using a learning signal in the communication signal; acquiring a detection target signal in the communication signal; an interference detection step of detecting interference in the detection target signal; a severity calculation step of calculating a severity of interference in the detection target signal by using the detection target signal, the interference detection result, and the learned signal state model; The interference detection method is performed by an interference detection device.

9. a signal state model learning step of learning a signal state model using a learning signal in the communication signal; acquiring a detection target signal in the communication signal; an interference detection step of detecting interference in the detection target signal; a severity calculation step of calculating a severity of interference in the detection target signal by using the detection target signal, the interference detection result, and the learned signal state model; A program for causing a computer to execute the above as an interference detection method.

Images